Mass Layoffs and Employment
How occupational employment is
affected by mass layoffs
An analysis of business establishment microdata—created by combining
microdata from the Occupational Employment Statistics program and
the Mass Layoff Statistics program—reveals that jobs lost between 2000
and 2007 in establishments where extended mass layoffs occurred tended
to be those which required less training and fewer analytical skills; jobs in
occupations that were core to the specific industry generally were retained
Dina Itkin
and
Laurie Salmon
Dina Itkin is an economist in the Office of
Occupational and Unemployment Statistics,
Bureau of Labor Statistics, Boston, MA. Email:
[email protected]. Laurie
Salmon is a supervisory
economist in the Office
of Occupational Employment Statistics, Bureau of
Labor Statistics, Washington, DC. Email: salmon.
[email protected].
I
n recent years, mass layoffs have affected
large numbers of workers.1 Even during times of stable employment levels
or economic expansion, mass layoffs occur
because of cost-cutting initiatives, relocation of operations, changes in technology
or consumer demand, or other reasons. Not
surprisingly, some occupations are more affected by these layoffs than are others. By
using a sample of establishments that had
at least one extended mass layoff during the
2000–2007 period, this article examines the
types of jobs affected by layoffs. An examination of this period offers insight into the
mass layoff effects on occupational employment before the start of the 2007–2009 recession.2
By combining data from two Bureau of
Labor Statistics programs—Mass Layoff
Statistics (MLS) and Occupational Employment Statistics (OES)—pre- and post-layoff
employment snapshots were compared for
each sampled establishment. The total employment of the 4,520 establishments in the
sample was more than 2.5 million before
layoffs and less than 2.2 million after layoffs—an overall decline in employment of
approximately 350,000 jobs, or 14 percent.
This study focuses on changes in occupational employment overall and by industry,
geographic region, and reason for the layoff.
The pattern of changes shows that, in
general, occupations that were retained or
whose employment expanded after layoffs
were those that tended to require analytical
skills and extensive technical training, such
as computer, financial, and legal analysts.
Establishments generally let go of workers
in occupations that tended to require less
training, such as clerical and personal care
occupations, and in occupations that tended
to require mainly nonanalytical skills, such
as material moving and production occupations. This finding was evident in the most
commonly reported reason for layoffs.
This overall pattern was driven, in part,
by the industries that experienced relatively
large numbers of layoff events and by the
occupations’ relative importance in their respective industries. Layoffs in the manufacturing and information technology industries during the study period contributed to
the employment declines in production and
computer occupations; however, these occupations’ relative importance to their respective industries seemed to lessen the impact
of the layoff.
A second finding was that establishments
were more likely to retain employment in
occupations that are core to their industry.
For example, establishments in finance and
insurance industries tended to increase employment in business and financial operations occupations, schools tended to increase
Monthly Labor Review • June 2011 3
Mass Layoffs and Employment
employment of teachers, and hospitals tended to increase
employment in healthcare occupations, despite layoffs in
other occupations. Other industries saw declines in core
occupations, but the declines were smaller than the declines in occupations with support functions.3 For example, manufacturing industries saw employment declines in
most occupational groups, but production workers were
laid off at lower rates than were production workers in
other industries.
This finding was most noticeable in geographic regions
where mass layoffs occurred in industries that were dominant in their economies. For example, the Midwest, which
had higher employment concentrations in manufacturing
and wholesale and retail trade industries than did other
regions, lost relatively fewer workers in occupations core
to those sectors: production, sales, and transportation occupations. The other regions had larger percentage declines in employment in these occupations. Likewise, the
West region, which had relatively high concentrations of
employment in motion picture and sound recording industries, lost relatively little employment in arts, design,
entertainment, and media occupations.
Methodology and data description
The first step was to identify establishments that had an
extended mass layoff and also had reported occupational
employment data to the OES survey both before and after
the layoff event. To obtain the largest number of sample
observations, establishments that had a layoff during the
2000–2007 period were matched with OES data from
1999 to 2008.
MLS defines the universe of extended mass layoffs as
private nonfarm establishments that had at least 50 initial
claims for unemployment insurance benefits filed against
them during a 5-week period, with at least 50 workers
separated for more than 30 days.4 Companies were identified by their State-specific unemployment insurance (UI)
number. From 2000 to 2007, there were 45,027 extended
mass layoff events. Excluding Puerto Rico, the full MLS
dataset contained 44,623 events. The OES data set comes
from a nationwide establishment survey of occupational
employment and wages. Some 400,000 business establishments are surveyed every year, and 2,611,373 distinct
establishments reported data between January 1999 and
May 2008.5
If multiple layoffs occurred within a company and county, only the first layoff event was used, in order to best
capture the occupations most vulnerable to layoff. Of the
44,623 layoff events, many were within both the same
4 Monthly Labor Review • June 2011
company and county. Counting only the first event, there
were 24,537 unique company/county layoff events. MLS
records were matched to OES establishments by State,
county, and UI account. There were 10,969 events that had
at least one corresponding OES observation. An MLS record may match more than one of the company’s establishments in the OES survey if there are several establishments within a county.
These “before” OES observations were linked to corresponding “after” observations by UI account number,
State, county, and Reporting Unit Number (RUN, a number that identifies a particular physical establishment under a parent company or UI account). The sample included
multiple matching pairs of OES observations, to maximize
the chance of capturing the establishments that had layoffs. The sample included all branches within the county
and UI account that had the same sets of physical establishments report both before and after the layoff. Hence,
the study essentially examines the effect of company-level
layoffs on staffing across branches within the county.
The study found 4,520 usable sets of “before layoff ” and
“after layoff ” OES observations, a total of 9,040 observations. The 4,520 observations that were in both the OES
and MLS data represented slightly more than 10 percent
of the layoff events during the 2000–2007 period. Because
total OES unweighted employment in these establishments before layoffs was 2,517,133, and total unweighted
employment after layoffs was 2,165,688, the net loss was
351,445 jobs.6
Approximately 96 percent of the OES observations occurred within the 4 years preceding the layoff event, and
nearly 90 percent of OES observations occurred within
the 4 years following the layoff event. Table 1 shows the
distribution of establishments by the number of years between the OES observation and the first layoff. The lag
Table 1. OES observations by number of years before and
after extended mass layoff
Number of
years
0
1
2
3
4
5
6
7
8
Total
Observations before layoff
715
1,342
1,303
754
232
101
55
17
1
4,520
Observations after
layoff
416
1,081
1,237
857
442
230
170
84
3
4,520
Cumulative
percentage
of total,
before
15.8
45.5
74.3
91.0
96.2
98.4
99.6
100.0
100.0
n/a
Cumulative
percentage
of total,
after
9.2
33.1
60.5
79.4
89.2
94.3
98.1
99.9
100.0
n/a
between data capture and layoff may impose limitations
on the research findings. By the time the establishment’s
staffing was captured the second time, employment may
have returned to its original levels, some workers may
have been re-hired, or staffing could have been influenced
by factors other than the layoff. There is also a chance that
an establishment underwent layoffs before the study’s
time window.
Another limitation is that MLS captures layoffs at the
company level within a county, whereas OES samples individual establishments within a company and county. Twothirds of the study units are known to be establishmentlevel matches because there exists only one establishment
in the company and county. For the other one-third of
study units, there is a chance that although MLS captured
a layoff event in a company in a particular county, the
matching OES units in the sample did not actually lay off
any workers. The layoff might have occurred elsewhere in
the company within the same county, but not necessarily
in the physical units surveyed.
A third limitation is that because this study requires OES
observations both before and after the layoff in order to
detect staffing changes, firms that go out of business completely before OES is able to sample them again are not
included. That is, the study includes only establishments
that go out of business if the permanent closure occurred
after the second OES observation. According to MLS statistics, there were 6,590 extended mass layoff events that
resulted in permanent worksite closures from 2000 to
2007. Because these closures were likely not random with
respect to each individual characteristic (e.g., industry, occupation, and region), there is attrition bias in the sample
selection.
About half the establishments in the sample had fewer
than 250 employees before layoffs. Table 2 shows the distribution of establishments by number of employees. After
layoffs, there was an aggregate shift toward smaller establishments. The number of establishments that were either
very small (1 to 9 employees), small (10 to 49 employees),
or medium-sized (50 to 249 employees) increased, while
the number of establishments with either a large (250 to
999 employees) or very large (more than 1,000 employees) size decreased.
Most establishments in the survey were in the West
and Midwest regions (defined later in this article). There
were also large numbers of establishments in the Southeast, Southwest, and New York-New Jersey regions. The
Mountain-Plains, Mid-Atlantic, and New England regions had less representation in the study sample. Table
2 also shows the distribution of establishments by region.
The industry sectors that had more than 100 establishments each were manufacturing; retail trade; construction; health care and social assistance; administrative and
support and waste management and remediation services;
information; transportation and warehousing; accommodation and food services; finance and insurance; and arts,
entertainment, and recreation. Table 2 shows the distribution of establishments by sector.
The primary reasons for layoffs that most establishments
in the sample reported were slack work/insufficient demand/nonseasonal slowdown, reorganization or restructuring, contract completion, financial difficulty, an extreme weather-related event, business-ownership change,
and contract cancellation. Table 3 shows the distribution
of establishments in the sample by primary layoff reason.
Finally, the study categorizes occupations into two
general groups: “analytical” and “nonanalytical.” This categorization is based on the 2000 Standard Occupational
Classification (SOC) system descriptions of tasks performed by each occupation and on Occupational Outlook
Handbook descriptions, and are supported by O*NET.7
The SOC definitions of detailed occupations in the analytical group often include words describing analysis. In
addition, occupations in the analytical group are related
to skills and abilities such as written expression, speaking, critical thinking, and deductive and inductive reasoning. At the major occupational group level, the analytical
group includes occupations such as legal, healthcare, and
business and financial operations. On the other hand, occupations in the nonanalytical group are related to skills
and abilities such as troubleshooting; repairing; dynamic,
explosive, static, and trunk strength; and stamina. The
nonanalytical group includes occupations such as production; transportation and material moving; office and administrative support; sales, installation, maintenance, and
repair; building and grounds cleaning and maintenance;
and protective service.
Employment changes by occupation
A comparison of employment before and after layoffs
shows that the largest numbers of jobs lost were in occupations that involved clerical or nonanalytical labor;
included among these jobs were those in the production,
office and administrative support, and transportation and
material moving occupational groups. Table 4 shows that,
despite layoffs, seven of the occupational groups grew,
including legal occupations; healthcare practitioners and
technical; healthcare support; and food preparation and
serving occupations. Employment in these occupations
Monthly Labor Review • June 2011 5
Mass Layoffs and Employment
Table 2. Establishments by industry sector, establishment size, and geographic region, sorted by number of establishments in
“before layoff” study sample
Category
Total number of establishments
Industry group
Number of
establishments in
full OES dataset,
1999–20081
Number of unique
company/county
layoff events in
full MLS dataset,
2000–2007
Number of
establishments
in "before layoff"
study sample
Number of
establishments
in "after layoff"
study sample
2,822,082
24,750
4,520
4,520
374,801
202,453
152,417
10,330
9,601
12,778
9,321
3,219
238
0
2,154
1,604
504
46
0
n/a
n/a
n/a
n/a
n/a
1,597,846
261,270
192,536
11,972
1,717
1,016
2,366
560
298
n/a
n/a
n/a
123,146
54,028
79,839
129,017
89,925
45,099
124,662
134,951
155,903
74,382
9,298
17,427
61,439
44,924
1,953
940
1,381
1,235
1,034
348
327
689
853
117
70
92
120
80
265
263
222
200
145
101
84
82
68
36
15
14
8
5
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
544,595
110,750
20,759
1,106,893
1,039,085
n/a
n/a
n/a
n/a
n/a
1,835
1,701
505
400
79
1,893
1,474
420
547
186
454,996
539,060
568,758
353,174
210,890
213,473
283,567
198,164
5,696
5,680
4,262
2,264
2,730
1,078
1,849
1,191
1,307
1,211
702
492
436
251
91
30
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
2
Goods-producing industries
NAICS 31–33 Manufacturing
NAICS 23 Construction
NAICS 21 Mining
NAICS 11 Farming
Service-providing industries
NAICS 44–45 Retail trade
NAICS 62 Health care and social assistance
NAICS 56 Administrative and support and waste management
and remediation services
NAICS 51 Information
NAICS 48–49 Transportation and warehousing
NAICS 72 Accommodation and food service
NAICS 52 Finance and insurance
NAICS 71 Arts, entertainment, and recreation
NAICS 81 Other services
NAICS 42 Wholesale trade
NAICS 54 Professional, scientific, and technical services
NAICS 61 Educational services
NAICS 22 Utilities
NAICS 55 Management of companies and enterprises
NAICS 53 Real estate and rental and leasing
NAICS 99 Public administration
Establishment size3
Establishments with 50–249 employees
Establishments with 250–999 employees
Establishments with 1,000 or more employees
Establishments with 10–49 employees
Establishments with 1–9 employees
Geographic region
West
Midwest
Southeast
Southwest
New York-New Jersey
Mountain-Plains
Mid-Atlantic
New England
Excluding Puerto Rico, Virgin Island, Guam
There are 849,435 establishments that have only an SIC code and no
NAICS code; these establishments are not included in the NAICS data and
so the sum of establishments by industry does not equal the total shown.
None of these 849,435 establishments were part of the MLS dataset.
3
MLS data do not include establishment size. Although the data include
the number of separations, they are not comparable because not all
establishments lay off the same proportion of employees.
increased 18 percent, 11 percent, 8 percent, and 4 percent, respectively. The occupational groups that grew were
service-providing occupations and, with the exception of
food preparation and serving occupations, tended to include higher paying and higher skilled occupations.
Even within the groups with the most job losses, detailed occupations that tended to have workers with more
training and education were least likely to experience layoffs. For instance, the detailed office support occupations
whose employment shrank the most tended to have more
workers whose educational attainment was a high school
diploma and short-term on-the-job training.8 Detailed
office support occupations with the greatest losses included customer service representatives; general office clerks;
1
2
6 Monthly Labor Review • June 2011
Table 3. Establishments by primary reason for extended mass layoff, 2000–2007
Reason for layoff
Number of unique company/
county layoff events in full MLS
dataset, 2000–2007
Number of establishments in
"before layoffs" study sample
Total number of establishments
24,750
4,520
Economic difficulties
Slack work/insufficient demand/non-seasonal business slowdown
Reorganization or restructuring of company
Contract completion
Financial difficulty
Bankruptcy
Business-ownership change
Contract cancellation
Extreme weather-related event
Import competition
Domestic relocation
Overseas relocation
Product line discontinued
Cost control/cost cutting/increase profitability
Labor dispute/contract negotiations/strike
Plant or machine repair/maintenance
Material or supply shortage
Automation/technological advances
Model changeover
Excess inventory/saturated market
Non-natural disaster
Energy related
Hazardous work environment
Natural disaster (not weather related)
Governmental regulations/intervention
Domestic competition
16,700
4195
3021
2424
1966
1003
940
726
543
487
313
188
187
157
143
96
63
39
39
35
32
30
23
23
14
13
3,078
926
757
346
277
39
116
115
130
74
39
26
49
52
36
18
16
15
13
4
8
3
7
8
0
4
Seasonal reasons
Seasonal
Other seasonal
Vacation period/school related or otherwise
6,072
4551
916
605
1,304
967
184
153
Other reasons
Data not provided: refusal
Data not provided: does not know
1,978
1061
917
138
60
78
shipping and traffic clerks; first-line supervisors; secretaries (except legal, medical, and executive); bill and account
collectors; and data entry keyers. Table 5 shows the 20
occupations with the largest declines in employment levels and table 6 shows those with the largest percent declines. For many of these occupations, most workers had
an educational attainment level of high school diploma or
equivalent.9
Among office support occupations that grew the most
following layoffs were interviewers, medical secretaries,
and payroll and timekeeping clerks. Table 7 shows the
20 occupations with the largest increases in employment
after layoffs, and table 8 shows those with the largest percent increases. More workers in these occupations had the
educational attainment of either a bachelor’s degree, or
some college or no degree.
Within the production occupations group, detailed occu-
pations with the most losses were team assemblers, miscellaneous metal and plastic workers, electrical and electronic
equipment assemblers, slaughterers and meat packers, and
first-line supervisors. For these types of assembly and fabrication jobs, a high school diploma was the most prevalent
level of education, but experience and additional training
were often needed for advanced assembly work.10 Employment declines due to productivity growth and strong foreign competition in manufacturing11 may have contributed
to the job losses evident in the layoff study sample. In fact,
a control group (to be explained in the next section) shows
that employment in production occupations declined by 2
percent among establishments from 2004 to 2008 regardless of layoff status, while employment in production occupations declined by 20 percent during a similar period in
establishments with mass layoffs.
Similarly, transportation occupations that lost the most
Monthly Labor Review • June 2011 7
Mass Layoffs and Employment
Table 4. Employment before and after extended mass layoff, by occupational group, 1999–2008, sorted from largest loss to
largest gain
Occupational group
Production
Office and administrative support
Transportation and material moving
Sales and related
Management
Installation, maintenance, and repair
Architecture and engineering
Personal care and service
Construction and extraction
Arts, design, entertainment, sports, and media
Protective service
Computer and mathematical science
Building and grounds cleaning and maintenance
Life, physical, and social science
Community and social services
Business and financial operations
Legal
Education, training, and library
Healthcare support
Food preparation and serving related
Healthcare practitioners and technical
Before layoffs
560,997
379,743
227,004
146,752
116,128
131,028
125,699
121,066
131,891
34,291
26,682
102,263
62,563
21,970
7,782
111,618
3,532
30,352
19,209
81,386
72,433
employment were those which involved predominantly
nonanalytical skills and short-term on-the-job training:
hand packers and packagers; freight, stock, and material
hand movers; and industrial truck drivers. Conversely,
employment increased in transportation occupations for
jobs that required either more training, or certification
or licensure: driver/sales workers,12 as well as excavating
and loading machine and dragline operators (with moderate-term on-the-job training being the most significant
source of training for the two kinds of operators).
Although employment in the computer occupational
group declined overall, employment grew in some of the
most highly skilled occupations in the group: computer
applications software engineers (which also saw the thirdhighest growth of all detailed occupations across groups)
and computer and network systems analysts. Employment declined, however, among computer programmers
and computer support specialists—occupations with job
functions that, according to SOC definitions, involve less
research and analysis.
Among business and financial operations occupations,
employment increased in those which involved analysis
and technical skills: management analysts, logisticians,
accountants, financial analysts, and personal financial advisors. The most significant source of education for the
five specified occupations was a bachelor’s degree, and
several of the occupations had high proportions of workers at the highest educational attainment level (doctoral
8 Monthly Labor Review • June 2011
After layoffs
441,624
307,211
186,961
117,806
94,388
111,526
106,762
105,212
121,397
28,389
21,232
97,675
60,547
20,997
7,908
111,847
4,161
31,429
20,800
84,849
80,353
Change after
layoffs
–119,373
–72,532
–40,043
–28,946
–21,740
–19,502
–18,937
–15,854
–10,494
–5,902
–5,450
–4,588
–2,016
–973
126
229
629
1,077
1,591
3,463
7,920
Percent change
after layoffs
–21.3
–19.1
–17.6
–19.7
–18.7
–14.9
–15.1
–13.1
–8.0
–17.2
–20.4
–4.5
–3.2
–4.4
1.6
0.2
17.8
3.5
8.3
4.3
10.9
or professional degree). Conversely, most employment
losses in this group were among occupations that generally required less academic preparation—training and
development specialists, buyers, and cost estimators. The
most significant source of education for buyers was longterm on-the-job training, and none of the aforementioned
three occupations had high proportions of workers who
had attained the highest educational level.13
The net number of jobs lost in each occupational group
is a result of the rates at which companies laid off different types of workers, as well as the types of workers
that tended to be employed in companies that had layoffs. To isolate these factors, the percent change is useful for assessing employment growth and decline relative
to an occupation’s initial employment level. (See tables 6
and 8.) The occupational groups that had the largest percent declines in employment were production, protective
service, sales and related occupations, office and administrative support, and management. The groups with the
highest percent growth in firms with mass layoffs were
legal occupations and healthcare practitioner occupations.
Healthcare practitioners; food preparation and serving;
healthcare support; and education, training, and library
occupations were the occupational groups with the highest levels of growth in employment. Some of the detailed
occupations that grew were service-related occupations
and included registered nurses, waiters and waitresses, cashiers, and interviewers. (See table 7.)
Table 5. The 20 occupations1 with the largest decline in employment level after extended mass layoff, 1999–2008, sorted by
size of decline
Employment
SOC
51–2092
41–2031
43–4051
Occupation
Before
layoffs
After
layoffs
Change
Percent
change
Number Number Percent
reporting reporting change in
before
after
number
layoffs
layoffs reporting
496
507
2.2
491
508
3.5
Production
Sales and related
Office and administrative
support
85,058
45,972
63,870
34,116
–21,188
–11,856
–24.9
–25.8
55,650
43,832
–11,818
–21.2
1394
1429
2.5
Production
16,930
6,341
–10,589
–62.5
105
80
–23.8
Production
22,914
14,684
–8,230
–35.9
186
177
–4.8
Production
Transportation and
material moving
14,358
6,149
–8,209
–57.2
30
21
–30.0
23,890
15,736
–8,154
–34.1
596
491
–17.6
Production
30,617
22,657
–7,960
–26.0
1678
1578
–6.0
Personal care and service
Office and administrative
support
Computer and mathematical science
13,936
6,508
–7,428
–53.3
77
90
16.9
29,746
22,661
–7,085
–23.8
2004
1921
–4.1
13,857
7,183
–6,674
–48.2
697
528
–24.2
Production
31,934 25,396
Production
25,968 19,498
Transportation and
material moving
52,270 45,992
43–5071
Office and administrative
support
20,581 14,607
41–9041
Sales and related
13,182
7,586
49–9042
Installation, maintenance,
and repair
32,387 26,801
17–2199
Architecture and
engineering
18,085 12,579
53–7051
Industrial truck and tractor
Transportation and
operators
material moving
24,398 18,943
11–9199
Managers, all other
Management
15,221
9,852
1
Excluded are any occupations with fewer than 10 reporting establishments before layoffs.
–6,538
–6,470
–20.5
–24.9
1170
375
1148
309
–1.9
–17.6
–6,278
–12.0
1269
1244
–2.0
–5,974
–5,596
–29.0
–42.5
1678
117
1582
113
–5.7
–3.4
–5,586
–17.2
1831
1907
4.2
–5,506
–30.4
405
311
–23.2
–5,455
–5,369
–22.4
–35.3
919
914
858
739
–6.6
–19.1
51–4199
51–2022
51–3023
53–7064
51–1011
Team assemblers
Retail salespersons
Customer service
representatives
Metal workers and plastic
workers, all other
Electrical and electronic
equipment assemblers
Slaughterers and meat
packers
Packers and packagers, hand
Occupational group
Establishments
43–9061
First-line supervisors/managers of production and
operating workers
Amusement and recreation
attendants
Office clerks, general
15–1021
Computer programmers
51–9061
Inspectors, testers, sorters,
samplers, and weighers
Production workers, all other
Laborers and freight, stock,
and material movers, hand
Shipping, receiving, and
traffic clerks
Telemarketers
Maintenance and repair
workers, general
Engineers, all other
39–3091
51–9199
53–7062
Comparison with a control group. A control group serves
to compare staffing changes among establishments that
experienced layoffs with occupational changes in the
economy as a whole. The change in published OES estimated employment from May 2004 to May 2008 was used
as the control group. The May 2004 estimates are based
on employment staffing patterns from November 2001 to
May 2004. Likewise, the May 2008 estimates are based
on staffing patterns from November 2005 to May 2008.
These periods cover a large portion of the study sample
frame. The employment changes for establishments that
had layoffs and for the economy as a whole are shown in
chart 1. The distance and direction from the 45-degree
line show the differences in behavior between establish-
ments with layoffs and the economy as a whole. Legal occupations make up the only occupational group above the
45-degree line; the group is the only one that grew more
in establishments with layoffs than in the control group
Quadrant I comprises occupational groups with employment growth in establishments that had layoffs (the
study group) and in the economy as a whole (the control group). Groups that grew in employment in both the
study and control subsets included healthcare and legal
occupations; food preparation and serving; and education,
training, and library occupations.
Occupational groups whose employment shrank in the
study subset but grew in the control subset are shown in
quadrant II. These groups were the most vulnerable to
Monthly Labor Review • June 2011 9
Mass Layoffs and Employment
Table 6. The 20 occupations1 with the largest percent decline in employment after extended mass layoff, 1999–2008, sorted by
size of decline
Employment
SOC
Occupation
53-7072
Pump operators, except wellhead
pumpers
Fire fighters
Communications equipment operators, all other
Mathematical scientists, all other
Occupational group
Establishments
Number Number Percent
Percent reporting reporting change in
change
before
after
number
layoffs
layoffs reporting
Before
layoffs
After
layoffs
Change
511
2,850
36
240
-475
-2,610
-93.0
-91.6
19
16
7
11
-63.2
-31.3
615
57
-558
-90.7
33
8
-75.8
365
37
-328
-89.9
17
10
-41.2
3,933
427
-3,506
-89.1
17
15
-11.8
888
112
-776
-87.4
24
14
-41.7
276
621
36
91
-240
-530
-87.0
-85.3
15
24
5
12
-66.7
-50.0
2,140
323
-1,817
-84.9
26
14
-46.2
1,438
273
-1,165
-81.0
30
32
6.7
Production
431
106
Office and administrative support
5,560
1,407
51-4194
Production
1,253
321
17-3021
Architecture and
engineering
2,618
735
51-2021
Production
1,490
432
29-1199
Healthcare practitioners
and technical
1,689
503
27-1027
Arts, design, entertainment, sports, and
media
156
47
51-7031 Model makers, wood
Production
318
96
49-9045 Refractory materials repairers,
Installation, mainteexcept brickmasons
nance, and repair
81
25
27-4014 Sound engineering technicians
Arts, design, entertainment, sports, and
media
277
86
1
Excluded are any occupations with fewer than 10 reporting establishments before layoffs.
-325
-75.4
13
10
-23.1
-4,153
-932
-74.7
-74.4
325
125
251
64
-22.8
-48.8
-1,883
-1,058
-71.9
-71.0
30
35
21
23
-30.0
-34.3
-1,186
-70.2
43
22
-48.8
-109
-222
-69.9
-69.8
16
25
12
16
-25.0
-36.0
-56
-69.1
10
6
-40.0
-191
-69.0
30
19
-36.7
33-2011
43-2099
15-2099
51-6091
51-8012
Extruding and forming machine
setters, operators, and tenders,
synthetic and glass fibers
Electrical and electronics repairers,
powerhouse, substation, and relay
Power distributors and dispatchers
51-7099
39-9021
Woodworkers, all other
Personal and home care aides
53-2012
Commercial pilots
51-2093
Timing device assemblers,
adjusters, and calibrators
Weighers, measurers, checkers, and
samplers, recordkeeping
Tool grinders, filers, and sharpeners
Aerospace engineering and operations technicians
Coil winders, tapers, and finishers
Health diagnosing and treating
practitioners, all other
Set and exhibit designers
49-2095
43-5111
Transportation and
material moving
Protective service
Office and administrative support
Computer and
mathematical science
Production
Installation, maintenance, and repair
Production
Production
Personal care and
service
Transportation and
material moving
layoff in struggling businesses despite overall growth elsewhere in the economy. The occupations with the greatest inverse relationship were arts, design, entertainment,
sports, and media; sales; and protective service occupations. Quadrant III shows occupational groups whose
employment shrank in both the control and study groups:
production, transportation, and material moving; and
management occupations.14
Finally, for comparison with another type of control,
table 2 shows the distribution of the full OES dataset from
1999 to May 2008 by region, industry sector, and establishment size. Table 3 shows the distribution of the full
MLS dataset of unique company/county layoff events from
2000 to 2007 by reason for layoff. A comparison shows
10 Monthly Labor Review • June 2011
that the sample is representative of the full data set, but
there are some exceptions. For example, manufacturing
had more representation in the study sample than in the
control group, which might explain why we see such large
employment changes in production occupations.
Regression analysis. Comparing the employment change
in the study group with published estimates of employment change is useful in assessing whether the study’s
results are reflected in the economy overall. Formal regression analysis achieves the same goal but also lets us
empirically control for other factors, such as industry,
geographic region, establishment size, and time between
observations.
Table 7. The 20 occupations1 with the largest increase in employment level after extended mass layoff, 1999–2008, sorted by
size of increase
Employment
SOC
Occupation
29–1111
Registered nurses
51–2031
Engine and other machine
assemblers
Computer software engineers,
applications
Assemblers and fabricators, all
other
Waiters and waitresses
15–1031
51–2099
35–3031
41–2011
43–4111
13–1111
35–3022
17–2072
51–9023
13–1081
41–3099
31–1012
49–2022
Cashiers
Interviewers, except eligibility and
loan
Management analysts
Counter attendants, cafeteria, food
concession, and coffee shop
Electronics engineers, except
computer
Mixing and blending machine
setters, operators, and tenders
Logisticians
Sales representatives, services, all
other
Nursing aides, orderlies, and
attendants
Telecommunications equipment
installers and repairers, except
line installers
Industrial machinery mechanics
Occupational group
Establishments
Number Number Percent
Percent reporting reporting change in
change
before
after
number
layoffs
layoffs reporting
Before
layoffs
After
layoffs
Change
34,917
40,876
5,959
17.1
321
291
–9.3
2,925
8,015
5,090
174.0
46
37
–19.6
16,704
21,679
4,975
29.8
393
447
13.7
Production
Food preparation and
serving related
Sales and related
Office and administrative support
Business and financial
operations
Food preparation and
serving related
Architecture and engineering
42,913
46,901
3,988
9.3
257
190
–26.1
18,934
23,293
22,356
26,514
3,422
3,221
18.1
13.8
211
669
200
655
–5.2
–2.1
4,294
7,057
2,763
64.3
79
90
13.9
8,645
11,408
2,763
32.0
461
568
23.2
4,183
6,928
2,745
65.6
174
155
–10.9
6,205
8,779
2,574
41.5
205
224
9.3
Production
Business and financial
operations
3,293
5,558
2,265
68.8
188
172
–8.5
1,857
4,098
2,241
120.7
219
342
56.2
Sales and related
6,002
7,879
1,877
31.3
214
458
114.0
Healthcare support
9,278
11,134
1,856
20.0
78
82
5.1
1,810
37.0
141
158
12.1
1,685
16.4
484
558
15.3
1,681
110.4
37
36
–2.7
1,678
71.4
30
34
13.3
1,661
64.2
517
889
72.0
1,636
34.3
387
477
23.3
Healthcare practitioners
and technical
Production
Computer and mathematical science
Installation, maintenance, and repair
4,898
6,708
49–9041
Installation, maintenance, and repair
10,244
11,929
19–1042 Medical scientists, except
Life, physical, and social
epidemiologists
science
1,522
3,203
47–2081 Drywall and ceiling tile installers
Construction and
extraction
2,349
4,027
13–1079 Human resources, training, and la- Business and financial
bor relations specialists, all other
operations
2,586
4,247
19–3021 Market research analysts
Life, physical, and social
science
4,769
6,405
1
Excluded are any occupations with fewer than 10 reporting establishments before layoffs.
Two sets of OES observations were created to run a regression. The study group was the set of 4,520 establishments that had layoffs. The control group was the set of
205,339 establishments that reported twice to the OES
survey in the study period—once in the 1999–2000 period and then again between November 2005 and May
2008—that was not in the study group.
There were 206,377 establishments that reported to OES
in both sets. Approximately 1,000 establishments that had
been in both the control group and the study group were
deleted from the control group to prevent duplication,
resulting in 205,339 remaining control observations. For
each pair of matching establishments, the change in employment by occupational group was calculated. The other
variables for this data set were region, goods-producing or
service-providing industry groups, establishment size, and
time between observations. The regression was based on a
total of 209,859 records.
The econometric model used was
∆employmentSOC major group = β0 + β1layoffi +
β2goods + β3totalemp_firsti + ΣjδjI(geographic regionji)
+ ΣjγjI(number of years between observationsji) + εi,
Monthly Labor Review • June 2011 11
Mass Layoffs and Employment
Table 8. The 20 occupations1 with the largest percent increase in employment after extended mass layoff, 1999–2008, sorted
by size of increase
Employment
SOC
Occupation
Occupational group
Establishments
Number Number Percent
Percent reporting reporting change in
change
before
after
number
layoffs
layoffs reporting
Before
layoffs
After
layoffs
98
412
314
320.4
15
15
0.0
60
78
191
237
131
159
218.3
203.8
19
13
44
12
131.6
-7.7
Production
Life, physical, and social
science
139
405
266
191.4
20
32
60.0
179
509
330
184.4
14
16
14.3
Production
Healthcare practitioners
and technical
Life, physical, and social
science
Food preparation and
serving related
2,925
8,015
5,090
174.0
46
37
-19.6
43
114
71
165.1
13
13
0.0
194
506
312
160.8
42
40
-4.8
67
174
107
159.7
11
24
118.2
Construction and
extraction
159
395
236
148.4
12
14
16.7
Personal care and service
933
2,239
1,306
140.0
29
31
6.9
61
141
80
131.1
16
23
43.8
131
300
169
129.0
12
19
58.3
438
992
554
126.5
180
293
62.8
175
394
219
125.1
37
67
81.1
142
316
174
122.5
44
85
93.2
1,857
4,098
Medical scientists, except
epidemiologists
1,522
3,203
49-2098
Security and fire alarm systems
installers
271
563
11-3049
Human resources managers,
all other
Management
909
1,851
1
Excluded are any occupations with fewer than 10 reporting establishments before layoffs.
2,241
120.7
219
342
56.2
1,681
110.4
37
36
-2.7
292
107.7
20
13
-35.0
942
103.6
331
637
92.4
43-4061
19-1012
39-9041
51-9193
19-2012
51-2031
29-1129
19-4011
35-2019
47-3011
39-6032
11-9039
47-3014
29-9011
27-1025
29-9012
13-1081
Eligibility interviewers,
government programs
Food scientists and
technologists
Residential advisors
Cooling and freezing equipment operators and tenders
Physicists
Engine and other machine
assemblers
Therapists, all other
Agricultural and food science
technicians
Cooks, all other
Helpers—brickmasons, blockmasons, stonemasons, and tile
and marble setters
Transportation attendants,
except flight attendants and
baggage porters
Education administrators, all
other
Helpers—painters, paperhangers, plasterers, and stucco
masons
Occupational health and safety
specialists
Interior designers
Occupational health and safety
technicians
Logisticians
19-1042
Office and administrative
support
Life, physical, and social
science
Personal care and service
Management
Construction and
extraction
Healthcare practitioners
and technical
Arts, design, entertainment, sports, and media
Healthcare practitioners
and technical
Business and financial
operations
Life, physical, and social
science
Installation, maintenance,
and repair
where the dependent variable was the change in employment level for each major occupational group, calculated
with the use of the first and second employment levels
reported for each occupational group. Layoff was an indicator dummy variable for whether the establishment had
a layoff; where layoffi = 1, a layoff occurred. Goods was a
dummy variable for the combined goods-producing industries, as opposed to the service-providing industries.
Totalemp_first was the total employment at the time of the
first observation of the establishment. The geographic region dummy variables indicated geographic regions: West,
12 Monthly Labor Review • June 2011
Change
Southwest, Southeast, Mountain-Plains, New York-New
Jersey, Midwest, Mid-Atlantic, and New England (captured in the intercept). Finally, there were nine dummy
variables representing the number of years between observations, ranging from 1 to 9; 9 years was captured in
the intercept.
In each of the 21 regressions, the significance of the
layoff coefficient indicates whether the establishments
in the study group were statistically different from establishments in the control group after controlling for these
other variables. For example, some of the employment
Chart 1. Change in employment after extended mass layoffs in study control groups
Percent change for
study group, 1999–2009
20
Quadrant IV
Percent change for
study group, 1999–2009
20
Quadrant I
Legal
15
15
Healthcare
practitioners and
technical
Food preparation and
serving related
10
5
0
Building and grounds
cleaning and maintenance
–5
Computer and mathematical science
5
Business and financial
operations
Community and
social services
Education, training,
and library
10
Healthcare support
Life, physical, and
social science
0
–5
Construction and extraction
–10
Installation,
maintenance, and repair
Transportation and
material moving
–15
–10
Personal care and service
Architecture and
engineering
–20
–25
Sales and related
Office and
Protective
administrative support service
Production
Quadrant III
–5
0
5
–15
Arts, design, entertainment,
sports, and media
Management
Quadrant II
10
15
–20
–25
20
Percent change for control group (total establishments), May 2004–May 2008
change in production occupations in the study group may
have been affected by the manufacturing plants that were
letting production workers go regardless of whether the
plant had mass layoffs. Appendix table A-1 shows the
output for these regressions.
Of the 21 layoff indicator coefficients, 14 were statistically different from zero at the 5-percent significance
level. Appendix table A-1 shows the layoff variable coefficients and their p-values. Because the dependent variable
is the change in employment level, the coefficients of the
variables are interpreted as the additional change in the
employment of an occupation because of layoff.
Controlling for the preceding variables, the model indicated that an extended mass layoff was associated with
a decline in employment for the following occupational
groups: production; office and administrative support;
sales and related; management; transportation and material moving; architecture and engineering; installation,
maintenance, and repair; construction and extraction;
personal care and service; and arts, design, entertainment,
sports, and media. An extended mass layoff was associated
with an additional employment decline of 20.8 percent in
production occupations, on average.
Conversely, with the same variables controlled for, an extended mass layoff was associated with growth in employ-
ment for the following occupational groups (presented in
descending order of magnitude of growth): food preparation and serving related; building and grounds cleaning
and maintenance; business and financial operations; and
legal.
These results are consistent with those set forth in the
previous section, with some exceptions. The occupational
groups with the largest declines in employment using regression also shrank relative to the control group in the
first comparisons. That is, the occupations with the largest
declines in the regression were generally in the lower left
area of chart 1 (quadrant III and part of quadrant II); this
is where declines were large in the group with layoffs relative to the economy as a whole. For the most part, occupational groups with differences that were not significant
in the regression comparison were closest to the 45 degree
line in the chart.
The largest differences between outcomes were in building and grounds cleaning and maintenance occupations
and food preparation and serving related occupations. In
the initial comparison, building and grounds cleaning and
maintenance occupations grew in the economy and shrank
in the layoff group. The regression comparison indicates
that building and grounds cleaning and maintenances occupations grew more in the layoff group relative to the
Monthly Labor Review • June 2011 13
Mass Layoffs and Employment
control group. In both comparisons, food service occupations grew in the layoff group and in the control group. In
the initial comparison to the economy as a whole, food
service occupations grew less in the layoff group than in
the control group; in the regression comparison to a control group, the food service occupations grew more in the
layoff group than in the control group.
Differences in the comparison may in part be due to the
differences in the control groups. The first control group,
where the control is the entire economy, captures growth
as a result of new establishments and may include a better
representation of smaller establishments. The second control group, which matches existing establishments at two
points in time, does not capture any “births,” or new establishments, in the comparison period. Also, because the
OES survey uses a probability-proportional-to-size sample, it is less likely that the matched set includes smaller
establishments.
Seasonal versus “economic difficulties” reasons for layoffs. Another regression analysis was conducted to determine
whether occupational changes differed significantly—after controlling for industry, region, time between observations, and establishment size—depending on whether the
layoff was due to seasonal reasons. On the basis of the
4,520 observations from the study sample, the primary
layoff reasons were grouped into three broad categories:
“economic difficulties,” “seasonal,” and “other.” The “economic difficulties” category included business demand,
financial difficulty, reorganization or restructuring of the
company, production, and domestic and overseas relocation reasons. The “other” category covered disaster/safety
reasons and unidentified reasons. The “seasonal” category
included seasonal, vacation period/school related or otherwise, and other seasonal reasons.15
The regression model used was
∆employmentSOC major group = β0 + β1layoffi × economici
+ β2 layoffi × seasonali + β3goods + β4totalemp_beforei +
ΣjδjI(geographic regionji) + ΣjγjI(number of years between
observationsji) + εi,
where the dependent variable was the change in employment level for each SOC major occupational group, calculated with the use of employment levels reported for each
occupational group in each establishment before and after
layoff. Layoff was a dummy variable indicating whether
the establishment had a layoff; where layoffi = 1, a layoff occurred. (In this data set, all observations had a lay14 Monthly Labor Review • June 2011
offi value of 1.) Economic, seasonal, and other were dummy
variables indicating the broad category of “reason for layoff.” Goods was a dummy variable for the combined goodsproducing industries, as opposed to the service-providing
industries. Totalemp_before was the total employment at
the time of the first observation of the establishment. The
geographic region dummy variables indicated geographic
region: New England (captured in the intercept), New
York-New Jersey, Mid-Atlantic, Southeast, Midwest,
Southwest, Mountain-Plains, and West. Finally, there
were nine dummy variables representing the number of
years between observations, ranging from 1 to 9; 9 years
was captured in the intercept.
Appendix table A-2 shows the output of these regressions. In none of the 21 regressions were the two reason
variables (economic and seasonal) statistically different from
each other at the 90-percent confidence level. This finding
suggests that, after other variables were controlled for, the
occupational changes did not differ significantly between
seasonal and economic layoff reasons. The data show that,
in the long term, establishments that had seasonal layoffs had staffing changes that were similar to establishments that had layoffs because of economic difficulties.16
It should be noted that some establishments that report
seasonal change as their primary reason for layoff might
also be undergoing other staffing changes. Because OES
surveys take place at the same time each year, changes as a
result of seasonal effects are mixed with other effects.
Occupations eliminated from establishments after layoffs. Another way to examine the effects of mass layoffs on jobs
in a particular occupation is to look at the change in the
number of establishments reporting employment in that
occupation after the layoff. This approach allows an examination of whether and how often establishments choose
to eliminate all workers in a certain occupation or, alternatively, choose to retain at least some employees in that
occupation.
The occupations whose employment count changed
from positive to zero were those which performed functions that businesses shed completely or outsourced after
layoffs. These occupations were predominantly auxiliary
administrative and managerial. The group whose occupations were most likely to be eliminated from establishments after layoffs was office and administrative support. Switchboard operators, including answering service,
topped the list, with 363 establishments eliminating the
occupation completely; the number of establishments reporting them dropped from 781 to 418. Several human
resources occupations had the same fate: employment,
recruitment, and placement specialists; training and development specialists; human resources assistants; and
payroll and timekeeping clerks. Other supporting administrative occupations affected were computer operators,
data entry keyers, file and procurement clerks, and janitors and cleaners.
Conversely, establishments reporting occupations commonly found in many businesses such as general managers and administrative clerks (bookkeeping, general office,
shipping, and payroll) tended to keep at least one of the
employees in those occupations. The share of establishments completely eliminating these occupations was relatively low, ranging from 1 percent to 12 percent. Workers
fulfilling these business functions apparently were considered essential for maintaining the basic operations of the
company.
Many of the occupations with the largest employment
losses overall were essential, or core, to their business, so
the occupations tended to be retained within the establishment, although at much lower levels of employment.
In fact, the three occupations with the largest employment declines—team assemblers, retail salespersons, and
customer service representatives—existed in more establishments after layoffs than before; the number of jobs
in the occupation, however, was smaller after the layoffs.
Similarly, although employment in sales occupations declined overall by almost 29,000 jobs, more establishments
reported having employment in sales occupations after
layoffs. This finding could be a result of shifts in staffing
patterns after restructuring. The effect of layoffs on employment in core occupations is discussed further in the
next section.
Occupational changes by industry sector
This section examines occupational employment changes
within and across industry groups. The first analysis shows
that, within sectors, core occupations generally were retained. Looking across sectors, the second analysis uses
regression to see how these changes differed between the
goods-producing and service-providing industry groups.
Employment changes within sectors. Examining employment changes by industry sector provides insight into the
effects of mass layoffs on the occupational structure of
specific types of businesses. It allows the identification of
core and support business functions in different industries
and shows that the severity of mass layoffs in terms of job
loss varied by occupation and the industry of the business
experiencing a layoff.
After layoffs, industry sectors that followed the pattern
of reducing employment in occupations requiring less
specialized skills and maintaining or increasing employment in analytical occupations included information; finance and insurance; professional, scientific, and technical
services; and the durable goods portion of the manufacturing sector. Examples of occupations with reduced employment in these sectors were sales and office workers;
examples of occupations with increased employment were
various types of analysts and engineers. These industry
sectors—and particularly the durable goods manufacturing portion of the manufacturing sector—experienced
large numbers of layoff events during the study period.
Establishments were more likely to retain employment
in occupations that were core to their industry. (See table
9.) For example, employment in business and financial
operations occupations grew in finance and insurance establishments with layoffs. The same was observed among
teachers in the education sector, as well as among healthcare workers in hospitals, extraction workers in mining,
and computer and mathematical science occupations in
the information sector. Other sectors saw smaller declines
in core occupations than in occupations that have support
functions. For example, in the manufacturing industries,
most occupational groups saw decreases, but production
workers had lower percent declines in manufacturing than
in several other industries. The same pattern was observed
in core occupations for other industries, including installation and maintenance occupations in the utilities sector,
transportation and material moving occupations in the
transportation sector, and personal care and service and
food preparation occupations in the accommodation and
food services sector.
The overall pattern of reducing the number of jobs in
occupations requiring less specialized skills and retaining
jobs in analytical occupations was driven by the industry
sectors with relatively large numbers of layoff events. These
sectors included information (NAICS 51); finance and insurance (NAICS 52); professional, scientific, and technical
services (NAICS 54); and the durable goods portion of the
manufacturing sector (NAICS 33). Employment declined
in these sectors for occupations requiring less specialized
skills, such as sales and office workers, while increasing for
various types of analysts and engineers.
In the fourth quarter of 2007, manufacturing industries
accounted for 24 percent of private nonfarm extended
layoff events. This study reflects that distribution, with
the three manufacturing components—food, wood, and
durable goods—experiencing the largest net losses in
employment compared to other industries. In durable
Monthly Labor Review • June 2011 15
Mass Layoffs and Employment
Table 9.
Percent change in employment after extended mass layoff, by (NAICS) industry and occupation, 1999–2008
Information
Goods-producing industries group
Occupational group
Management
Business and financial
operations
Computer and mathematical science
Architecture and
engineering
Life, physical, and social
science
Community and social
services
Legal
Education, training, and
library
Arts, design, entertainment, sports, and media
Healthcare practitioners
and technical
Healthcare support
Protective service
Food preparation and
serving related
Building and grounds
cleaning and maintenance
Personal care and service
Sales and related
Office and administrative
support
Construction and
extraction
Installation, maintenance,
and repair
Production
Transportation and material moving
Mining
(21)
ManuConfacturstruction
ing
(23)
(31)
Manufacturing
(32)
Financial
activities
Professional and business services
ProfesReal essional,
Finance tate and
ManuInforscienand in- rental
facturmation
tific, and
surance
and
ing
(51)
technical
(52)
leasing
(33)
services
(53)
(54)
ManageAdministrative
ment of
and support and
compawaste managenies and
ment and remeenterdiation services
prises
(56)
(55)
18.7
–3.4
–43.2
–11.5
–13.9
–51.0
2.4
–43.8
5.3
–59.2
–27.7
118.6
20.2
–29.2
–17.0
–1.7
6.6
2.1
2.5
59.8
–26.7
2.9
57.6
–.5
–43.4
.1
–10.6
13.8
11.6
–28.7
–16.1
3.3
–17.9
2.2
–14.0
–48.5
–32.7
–13.8
–19.6
29.8
.0
–14.1
67.6
–48.9
50.9
26.3
8.2
–15.3
–23.6
9.5
80.0
(1)
–2.1
(1)
8.0
.0
(1)
.0
211.1
(1)
–47.6
(1)
85.0
.0
25.5
(1)
13.4
–54.5
9.7
.0
(1)
(1)
–14.0
(1)
2.6
218.0
–3.8
.0
(1)
(1)
–100.0
–6.7
–71.9
94.1
.0
37.5
12.8
91.2
(1)
–54.0
–19.1
108.2
.4
–5.9
1.5
–67.7
–4.2
–1.0
–41.5
–11.7
.0
6.7
–20.6
(1)
–2.2
12.6
(1)
–55.0
91.0
.0
–50.8
–19.7
–50.0
–27.0
73.9
.0
–54.0
14.3
–23.3
–7.9
(1)
.0
–42.9
–70.0
(1)
.0
–17.5
(1)
–14.3
8.6
–64.9
–14.5
(1)
–27.1
216.8
–90.9
–63.0
–75.0
138.1
–24.7
(1)
–23.9
16.4
70.0
(1)
208.6
–18.4
–50.7
–31.6
–16.7
(1)
–24.0
–38.6
(1)
61.8
–42.4
–50.0
–16.5
–60.2
–55.4
–41.7
–48.9
(1)
–35.2
(1)
.0
–71.0
46.0
(1)
–5.9
.6
–7.8
–73.8
–1.5
30.1
–36.9
6.1
4.1
–21.2
–16.5
–21.6
–34.3
–33.2
31.6
–21.7
–21.2
–20.6
25.5
–2.8
–42.6
–35.9
–29.2
–57.3
(1)
(1)
–71.9
(1)
21.4
3.9
–30.4
7.4
42.2
–13.0
–23.4
–27.2
–20.2
–22.6
–22.8
8.1
–28.1
–26.8
61.2
.0
(1)
–60.4
–58.2
–22.3
–34.9
12.1
11.0
–12.8
.0
–17.0
–30.9
–22.1
–47.9
31.4
.2
3.4
–65.9
8.2
See note at end of table.
goods manufacturing—which had the largest net employment loss of the three manufacturing components
in the study—losses in employment levels were mostly in
production occupations, such as team assemblers, electrical equipment assemblers, and weighers. Durable goods
manufacturers hired workers in analytical occupations,
such as electronics engineers, computer applications software engineers, and logisticians.
Within durable goods manufacturing, transportation
16 Monthly Labor Review • June 2011
equipment manufacturing (NAICS 336) had the highest
net employment loss. Table 10 shows the transportation
equipment manufacturing occupations that shrank by
more than 1,000 jobs in the study group. Most of the
losses were in production occupations. Some occupations
that grew were related to product design and engineering:
computer software applications engineers, logisticians,
commercial and industrial designers, and mechanical
engineers.
Table 9. Continued—Percent change in employment after extended mass layoff, by (NAICS) industry and occupation, 1999–2008
Education and
health care
Trade, transportation, and utilities
Occupational group
Management
Business and financial
operations
Computer and mathematical science
Architecture and
engineering
Life, physical, and social
science
Community and social
services
Legal
Education, training, and
library
Arts, design, entertainment, sports, and media
Healthcare practitioners
and technical
Healthcare support
Protective service
Food preparation and
serving related
Building and grounds
cleaning and maintenance
Personal care and service
Sales and related
Office and administrative
support
Construction and
extraction
Installation, maintenance,
and repair
Production
Transportation and material moving
1
Other Public
seradminvices istration
Leisure and
hospitality
Health
EducaArts,
Accomcare and
Other Public
tional
entertain- modation
social
seradminserment, and and food
assisvices istration
vices
recreation services
tance
(81)
(99)
(61)
(71)
(72)
(62)
Utilities
(22)
Wholesale
trade
(42)
Retail
trade
(44)
Retail
trade
(45)
Transportation
(48)
Warehousing
(49)
–21.9
–28.4
–34.4
–17.0
–25.1
–26.0
–6.9
–7.7
–7.6
–25.1
–11.3
–3.5
–24.1
–22.7
1.4
8.6
–11.2
13.9
–2.3
6.4
48.2
9.9
25.4
–16.4
–37.3
–26.5
–1.8
–9.6
13.2
31.9
1.6
4.0
57.9
–2.4
86.4
–30.1
–74.4
21.8
–43.0
–55.8
–45.6
78.4
35.0
–46.1
409.4
–54.8
–4.4
–69.7
–94.5
–34.3
9.4
–40.4
–10.6
.0
84.5
–7.7
–8.8
–62.7
14.3
(1)
.0
20.8
.0
–31.5
.0
270.0
.0
81.3
(1)
–29.5
.0
(1)
–33.9
33.3
10.4
104.5
(1)
472.7
(1)
109.1
–29.3
(1)
.0
(1)
(1)
(1)
(1)
(1)
–46.0
.0
2.7
5.0
–23.2
–22.2
13.5
(1)
161.4
–19.2
21.1
34.4
–9.9
–90.0
12.6
22.9
–48.6
9.4
126.6
–42.8
(1)
.0
–91.9
–47.1
.0
–44.7
55.1
–38.0
–68.1
13.8
108.2
–12.3
–33.8
–19.4
–10.2
31.3
(1)
–34.1
–2.0
–5.0
–11.0
12.1
11.4
9.7
–10.6
–91.4
–35.9
93.6
48.7
–1.8
21.1
(1)
–42.0
(1)
.0
(1)
.0
–99.0
13.2
–30.4
24.4
(1)
15.4
.5
76.3
–6.5
23.5
(1)
–81.4
(1)
90.8
35.9
.0
–46.1
–49.0
–43.1
14.2
–22.2
–4.3
–21.6
13.0
–9.7
–14.4
–64.7
(1)
–19.0
–2.2
–59.2
7.6
8.7
–39.0
23.6
52.4
–24.9
–40.3
–5.0
–5.0
–13.9
–16.4
–.6
–32.5
–35.9
.0
2400.0
–40.5
–33.7
–10.7
–11.9
–21.3
42.9
2.1
6.7
–8.0
–5.2
–25.8
–45.3
85.5
50.3
–41.3
–80.4
32.3
(1)
–23.9
–37.0
–39.1
19.9
–34.2
50.9
–18.4
–71.6
12.5
–9.0
–9.7
–39.3
–16.4
–10.0
–15.5
8.3
–79.1
20.3
26.6
–43.2
31.4
–10.5
–11.3
43.8
–6.3
–12.9
–49.5
–22.7
–28.6
–30.9
–12.5
–25.5
–16.4
–19.0
–15.0
12.7
348.3
–40.6
–21.4
–9.5
–65.8
–53.9
Percent change excluded because it is based on fewer than 5 establishments reporting occupations in the occupational group before layoffs.
Establishments that had layoffs in the information sector reduced employment in occupations that require less
specialized training: sales supervisors and representatives,
customer service representatives, and stock clerks. After
layoffs, they had higher employment in occupations involving technical skills: computer software engineers;
telecommunications equipment repairers; management,
computer systems, and network systems analysts; and accountants and auditors.
Similarly, most finance and insurance businesses which
had layoffs shed jobs in occupations that tended to pay less
and that did not include analysis as a primary job function:
clerical workers, such as customer service representatives,
telemarketers, brokerage clerks, and general office clerks.
These establishments increased employment in analytical
occupations, such as computer systems analysts, financial
analysts, market research analysts, and management analysts. They also added technical positions that tended to
Monthly Labor Review • June 2011 17
Mass Layoffs and Employment
Table 10. Transportation equipment manufacturing (NAICS 336) occupations whose employment declined by at least 1,000
after extended mass layoff, 1999–2008
Employment
SOC
51–2092
51–4199
51–9199
51–1011
17–2011
51–2011
51–4121
51–4111
53–7051
51–9061
43–5061
49–9042
53–7062
51–9122
13–1199
53–6051
51–4031
Occupation
Team assemblers
Metal workers and plastic workers, all other
Production workers, all other
First-line supervisors/managers of production and
operating workers
Aerospace engineers
Aircraft structure, surfaces, rigging, and systems
assemblers
Welders, cutters, solderers, and brazers
Tool and die makers
Industrial truck and tractor operators
Inspectors, testers, sorters, samplers, and weighers
Production, planning, and expediting clerks
Maintenance and repair workers, general
Laborers and freight, stock, and material movers,
hand
Painters, transportation equipment
Business operations specialists, all other
Transportation inspectors
Cutting, punching, and press machine setters,
operators, and tenders, metal and plastic
–12,881
–8,140
–4,477
Number
reporting before layoffs
104
39
59
Number
reporting after
layoffs
93
42
55
5,681
5,360
–3,002
–2,906
228
24
205
18
6,801
6,720
6,629
5,090
9,020
3,060
5,623
4,228
4,493
4,651
3,195
7,357
1,469
4,079
–2,573
–2,227
–1,978
–1,895
–1,663
–1,591
–1,544
15
126
126
117
184
159
166
11
109
121
115
175
143
163
4,831
3,998
6,966
1,508
3,371
2,564
5,698
332
–1,460
–1,434
–1,268
–1,176
99
61
105
20
97
56
88
7
3,570
2,404
–1,166
74
59
Before layoffs
After layoffs
Change
36,788
13,208
10,973
23,907
5,068
6,496
8,683
8,266
be highly paid, such as computer software engineers, accountants and auditors, and personal financial advisors.
The professional, scientific, and technical services sector also followed this pattern. These businesses reduced
the employment of general office clerks, computer support specialists, customer service representatives, and data
entry keyers. They added computer systems analysts, management analysts, and market research analysts, in addition to accountants and auditors.
Most establishments in the health care and social assistance sector tended to lay off administrative support occupations not directly related to healthcare, such as general
office and billing clerks. They hired health care workers
including registered nurses, nursing aides, and licensed
practical nurses. The number of medical secretaries grew,
but by less than the employment decline among other administrative support occupations.
Four sectors fared relatively well after layoffs and grew
in total employment. Those with net gains in employment
were health care and social assistance (NAICS 62); educational services (NAICS 61); mining (NAICS 21); and postal
service/couriers/warehousing (NAICS 49). (See table 9.)
Within health care and social assistance—which was the
sector with the highest net gain in employment—hospitals and ambulatory health care services grew the most,
increasing the number of jobs with functions related to
health care and administration: office and administrative
18 Monthly Labor Review • June 2011
Establishments
support, business and financial operations, management,
and computer and mathematical science occupations.
Production occupations experienced the largest losses;
hospitals especially reduced the number of laundry and
drycleaning jobs.
It is informative to examine the occupations that declined in employment after layoffs in sectors which nonetheless experienced net employment gains. Personal and
home care aides lost the most employment overall in the
health care sector. Mining establishments that underwent layoffs shed several occupations that required less
specialized training and skills: general maintenance and
repair workers, industrial truck and tractor operators, and
machinery maintenance workers. They added operating
engineers, industrial machinery mechanics, heavy and
tractor-trailer truck drivers, and mobile heavy equipment
mechanics.
Occupational changes in goods-producing versus service-providing establishments. A regression analysis was conducted to analyze the effect of layoffs by goods-producing and
service-providing aggregations (simply termed “groups”)
on occupational employment, controlling for region, time
between observations, and establishment size. The variables of interest were the two interaction dummy variables
for layoff × group. To see if the large employment decline
in production occupations in goods-producing industries
was due to the overall decline in employment in production industries, the model also includes a non-interaction
dummy variable for the group of goods-producing establishments. The regression was based on a total of 209,858
observations—205,339 control observations and 4,520
study observations.
The model used was
∆employmentSOC major group = β0 + β1 layoffi × goodsproducing group + β2 layoffi × service-providing group
+ β3totalemp_firsti + β4goods-producing group +
ΣjδjI(geographic regionji) + ΣjγjI(number of years between
observationsji) + εi,
where the dependent variable was the change in employment level for each occupational group between the first
observation and the second. Layoff was a dummy variable
indicating whether the establishment had a layoff; where
layoffi = 1, a layoff occurred. Goods was a dummy variable
for the goods-producing aggregation, as opposed to the
service-providing aggregation. Layoff × goods and layoff ×
service were interaction dummy variables. Totalemp_first
was the total employment at the time of the first observation of the establishment. The geographic region dummy
variables indicated geographic region: West, Southwest,
Southeast, Mountain-Plains, New York-New Jersey,
Midwest, Mid-Atlantic, and New England (captured in
the intercept). Finally, there were nine dummy variables
representing the number of years between observations,
ranging from 1 to 9; 9 years was captured in the intercept.
With region, time between observations, and establishment size controlled for, an extended mass layoff in the
goods-producing aggregation was associated with a greater
employment decline than in the service-providing aggregation for two occupational groups: architecture and
engineering; and installation, maintenance and repair.
For production occupations, layoffs were associated with
employment decline in the goods-producing group, while
layoffs in the service-providing group were actually associated with slight employment growth (significant only at
the 11-percent level). The differences between the interaction term coefficients were statistically significant, and
the coefficients themselves were statistically significant.
Appendix table A-3 shows the regression parameter estimates and statistics for the goods-producing and serviceproviding interaction variables.
With region, time between observations, and establishment size controlled for, an extended mass layoff in the
service-providing aggregation was associated with a greater employment decline than in the aggregation of goods-
producing sectors for management and for sales and related occupations. For sales and office and administrative
occupations, the employment decline was substantially
greater in the service-providing group. In protective service occupations and personal care and service, a layoff in
the service-providing group was associated with employment decline, while a layoff in the goods-producing group
was associated with employment growth.
Conversely, in building and grounds cleaning and maintenance occupations, an extended mass layoff in the service-providing group was associated with greater employment growth than in the goods-producing group.
Finally, for three occupational groups, the individual
goods-producing and service-providing group parameter
estimates were significant, but the differences between the
two were not. Employment in transportation and material
moving occupations declined the same amount in both
groups. The difference between them was not statistically
significant. Similarly, employment in legal occupations
and food preparation and serving-related occupations
grew significantly in both the goods-producing and service-providing groups, but the difference between the two
coefficients was not statistically significant.
Occupational changes by reason for layoff
The MLS program asks employers for a primary reason for
the layoff. Employers could report 30 reasons for extended mass layoffs over the study period. These reasons can
be grouped into six broad categories: business demand,
financial, organizational, production, disaster/safety, and
seasonal. Business demand accounted for 34 percent of
the events in the fourth quarter of 2007, the highest in the
economic reasons category excluding seasonal and other
reasons. Extended mass layoffs stemming from financial
issues accounted for 7 percent of layoff events, the next
highest in the economic reasons category. (See table 11.)17
The pattern of employers retaining or adding workers
in higher skilled analytical or technical occupations while
letting go of workers in occupations that require nonanalytical or office and clerical skills was, in general, evident
regardless of the reason for the layoff. It is apparent from
table 11 that workers in production, material moving, installation and maintenance, and office and administrative
support occupations were let go after almost all types of
layoffs. The number of jobs in computer and mathematical science occupations, architecture and engineering occupations, and business and financial operations occupations either grew, or declined proportionally less than the
number of jobs for lesser skilled workers, regardless of the
Monthly Labor Review • June 2011 19
Mass Layoffs and Employment
Table 11. Percent change in employment in the study group after extended mass layoff, by primary reason for layoff and occupation, 1999–2008
Business demand
Occupational
group
Disaster/safety
Financial
Slack
work/in- HazardCost conExcess
Natural
Dosufficient
ous
Non- Extreme
trol/cost
inven- Import
FinanContract Contract
disaster
mestic
demand/ work
natural weather- Bank- cutting/
tory/sat- compecial difcancella- comple(not
compenonenvidisas- related ruptcy increase
urated
tition
ficulty
tion
tion
weather
tition
seasonal
ronter
event
profitmarket
related)
business ment
ability
slowdown
–21.4
1.3
–55.0
–24.3
–36.7
–22.0
9.3
–9.1
–32.9
13.1
–16.9
–39.5
–13.0
Management
Business and
financial operations
9.4
Computer and
mathematical
science
–29.8
Architecture and
engineering
–51.2
Life, physical, and
social science
–49.6
Community and
social services
–69.4
Legal
46.7
Education, training, and library
32.6
Arts, design,
entertainment,
sports, and
media
14.2
Healthcare practitioners and
technical
–37.2
Healthcare
support
(1)
Protective service
–16.7
Food preparation
and serving
related
–72.7
Building and
grounds
cleaning and
maintenance
–6.9
Personal care
and service
–31.4
Sales and related
–.5
Office and administrative support
–34.9
Construction and
extraction
19.8
Installation,
maintenance,
and repair
–14.5
Production
–16.1
Transportation
and material
moving
–12.1
See notes at end of table.
21.8
–78.6
–23.5
–10.6
–7.5
–21.4
–16.1
280.0
4.2
–27.0
.3
9.0
–10.1
(1)
(1)
–25.1
–12.4
5.0
77.8
(1)
5.9
–27.0
41.3
–20.3
23.5
–43.8
–44.4
–25.2
–26.2
–2.0
170.4
(1)
–7.2
–46.3
–76.1
–15.5
11.5
.0
.0
25.0
19.6
57.1
(1)
.0
–40.2
218.4
–36.8
–2.1
92.2
8.7
.0
.0
.0
.0
.0
(1)
7.4
10.8
.0
.0
.0
.0
.0
.0
107.4
(1)
(1)
–22.5
.0
11.4
10.8
.3
–14.8
.0
.0
(1)
–35.3
.0
.0
.0
281.3
(1)
(1)
2.5
–81.4
.0
.0
–51.3
–10.1
(1)
(1)
.0
111.8
–40.5
40.6
–12.0
27.3
(1)
(1)
–50.0
–7.3
(1)
(1)
(1)
–30.5
–7.0
32.5
23.3
(1)
–8.2
.0
.0
.0
.0
.0
–44.4
–42.5
–5.2
.0
(1)
.0
(1)
.0
–29.4
–47.8
–35.6
–1.3
–23.2
(1)
50.3
6.8
30.4
–16.3
.0
.0
(1)
0.6
–100.0
–3.3
–23.9
–23.4
–15.6
7.9
21.2
–6.2
(1)
(1)
–57.2
–4.5
(1)
–68.4
9.4
–40.4
–57.2
32.8
2.3
–43.0
–38.2
.0
.0
.0
.0
.0
–58.3
–16.3
–24.3
(1)
(1)
(1)
–32.4
–68.4
–16.1
–40.2
–29.1
–9.1
–29.3
50.3
–17.4
–11.6
–21.8
1.0
–55.6
–2.9
–36.8
–15.6
–36.7
–51.2
22.7
–27.4
–17.7
–27.7
–11.2
1.2
–42.0
(1)
–39.2
–13.8
–28.8
22.3
(1)
49.7
–56.6
88.2
3.7
–22.2
–9.1
–21.1
–66.1
–27.7
–7.5
–51.9
–50.4
–14.3
–17.8
–4.6
–28.2
–9.1
5.6
–20.5
–41.3
12.7
13.5
–27.0
–5.4
–26.6
–36.7
–14.0
–32.8
14.8
–69.4
–72.7
–52.5
–7.5
–46.1
–41.0
75.6
–31.0
–12.2
2.6
–34.4
reason for the layoff, although there were some exceptions.
At the detailed occupation level, employment of customer
service representatives, general office clerks, and bookkeeping clerks declined after most types of layoffs.
20 Monthly Labor Review • June 2011
This overall pattern was driven, in part, by layoffs due to
a number of reasons: the reorganization or restructuring
of a business, a change in ownership, financial difficulty,
slack work, competition from imports, cost control or cost
Table 11. Continued—Percent change in employment after extended mass layoff, by primary reason for layoff and
occupation, 1999–2008
Organizational
Occupational
group
Business–
ownership
change
Production
Reorganiza- AutoGovern- Labor
tion or mation/
mental dispute/
retechno- Energy regula- contract
struc- logical related tions/
negoturing
adinter- tiations/
of com- vances
vention
strike
pany
–12.5
–45.0
–36.8
0.0
21.1
Other
Material or
supply
shortage
Plant
or maProduct
DoModel
chine
Overseas
line
mestic
change- repair/
relocadiscon- relocaover
maintion
tinued
tion
tenance
Management
–37.3
Business and
financial operations
25.1
–4.0
–.7
–18.6
.0
Computer and
mathematical
.0
science
–17.1
9.3
57.7
(1)
Architecture and
engineering
–26.6
–12.7
111.6
(1)
.0
Life, physical, and
social science
–47.4
–11.3
–75.5
(1)
.0
Community and
social services
–14.0
–17.1
.0
.0
.0
Legal
12.4
14.7
(1)
.0
.0
Education, training, and library
–95.3
–31.4
.0
.0
.0
Arts, design,
entertainment,
sports, and
media
–55.7
46.1
–14.3
.0
.0
Healthcare practitioners and
technical
–17.5
5.9
–15.8
(1)
.0
Healthcare
support
–16.5
13.1
(1)
.0
.0
Protective service
–57.9
–19.9
–21.3
(1)
.0
Food preparation
and serving
related
–12.8
10.7
–19.7
.0
.0
Building and
grounds
cleaning and
maintenance
–25.4
–6.3
–57.2
(1)
.0
Personal care
and service
9.9
–8.4
33.8
–3.1
.0
Sales and related
–32.6
–23.3
–32.9
(1)
.0
Office and administrative support
–18.7
–23.4
–35.1
6.0
.0
Construction and
extraction
–92.7
–28.6
–47.6
(1)
.0
Installation,
maintenance,
and repair
–26.7
–14.2
–40.1
–4.4
.0
Production
–31.4
–29.3
160.7
–28.9
.0
Transportation
and material
moving
–43.5
–34.4
–57.3
–31.4
.0
NOTE: Table does not show "data not provided (refusal);" "data not
provided (does not know);" "seasonal;" "vacation period/school related
or otherwise;" or "other seasonal."
–19.0
–31.1
–13.0
–31.8
11.5
–60.1
8.9
–20.7
–30.9
16.2
146.5
40.6
–16.0
20.8
30.0
–45.5
5.9
112.7
2.4
–61.1
48.5
–7.2
–47.1
–3.8
36.3
14.0
–57.7
–61.7
–41.7
–73.3
–37.3
–56.4
–77.7
–88.5
.0
.0
.0
.0
.0
.0
.0
.0
.0
(1)
.0
(1)
.0
–46.2
–45.5
.0
.0
.0
(1)
(1)
(1)
–1.8
(1)
–100.0
(1)
33.0
111.5
54.2
39.0
(1)
–57.1
(1)
–16.3
–15.6
–84.6
48.7
–4.8
.0
(1)
(1)
(1)
.0
(1)
(1)
51.0
(1)
–95.7
(1)
–12.5
3.4
.0
24.7
–24.7
(1)
25.9
–100.0
–34.6
(1)
14.3
355.6
–26.4
–85.7
–15.9
–87.5
24.7
.0
(1)
(1)
–89.3
.0
4.7
.0
52.2
5.7
–2.2
.0
–55.1
–21.2
–61.1
–76.7
3.8
–6.1
–40.3
12.2
–70.4
2.1
–1.0
–70.0
–18.3
–93.8
–75.0
5.3
–26.4
–28.1
–13.4
13.6
–8.6
–7.7
1.9
–30.1
–29.2
–59.8
–68.8
–21.3
–36.7
–14.2
–25.2
176.3
–38.9
14.4
–56.9
–56.8
Percent change excluded because it is based on fewer than 5 establishments reporting occupations in the occupational group before
layoffs.
cutting, and the relocation of domestic work.18 These
reasons accounted for a large number of layoffs and
a large share of the employment losses among lesser
skilled occupations and increased employment in
analytical or technical occupations during the period
studied. Layoffs that occurred after either the reloca-
1
Monthly Labor Review • June 2011 21
Mass Layoffs and Employment
tion of domestic work or the discontinuation of a product
line followed the pattern closely. However, the types of
jobs affected by layoffs often depended on the reason for
the layoff. Patterns within each group are examined next.
Organizational change. The largest cause of job loss from
layoffs was organizational change, which includes the reorganization or restructuring of companies and changes
in business ownership. Establishments that reorganized
or restructured (representing 10 percent of all layoffs, the
third most commonly reported reason for a layoff event19)
tended to eliminate jobs in production occupations—these
jobs declined by more than 27,000—and in administrative
support occupations, such as customer service representatives, general office clerks, and data entry keyers; employment in administrative support occupations declined by
23,000. Businesses that reorganized or restructured also
reduced jobs for occupations that involved less technical
skill, such as retail salespersons, hand laborers, hand packers, team assemblers, and general maintenance workers.
Employers cut back on jobs in some technical occupations
and added jobs in others, resulting in a net gain in computer and mathematical occupations.
Reorganized and restructured establishments hired more
workers in occupations that develop new software and
applications—occupations such as computer systems and
applications software engineers, computer systems analysts, computer hardware engineers, engineering managers, electronics engineers, telecommunications equipment
installers, and logisticians—while reducing the number of
other technical jobs—such as jobs for computer programmers, who primarily code programs for existing software.
Some occupational groups, however, fared well after this
type of layoff: arts, design, entertainment, sports, and media; education; legal; community and social service; and
life, physical, and social sciences occupations.
Establishments with layoffs resulting from businessownership changes, the seventh most commonly reported
reason for a layoff event,20 followed the pattern of shedding workers in less technical occupations and hiring additional analytical workers. After production, office and
administrative, and transportation and material moving
occupations, sales workers accounted for most of the job
loss. Following layoffs induced by business-ownership
changes, establishments had fewer workers in occupations related to sales, marketing, and maintenance; these
occupations included customer service representatives,
general office clerks, marketing managers, market research analysts, sales managers, janitors and cleaners, and
maintenance and repair workers. The production occupa22 Monthly Labor Review • June 2011
tion that lost the most employment was textile cutting
machine setters, operators, and tenders. As was seen in
businesses that reorganized, the computer occupations
that declined in employment among establishments with
business-ownership changes required less technical skill
than those which increased in employment. Other occupations that grew involved financial and accounting business functions: among these occupations were payroll and
timekeeping clerks; management analysts; bookkeeping,
accounting, and auditing clerks; financial managers; and
accountants and auditors.
Business demand. The second-largest cause of job loss
from layoffs was a decline in business demand. Compared
with other reasons for layoffs, business demand factors
resulted in relatively greater losses of technical workers
and also resulted in large losses of lesser skilled workers.
Specifically, layoffs due to slack work, insufficient demand,
and nonseasonal business slowdown resulted in the largest employment declines among any of the 30 reasons for
layoffs. Production occupations accounted for the most
losses after this type of layoff. The production occupations that topped the list of losses were aircraft structure,
surfaces, rigging, and systems assemblers; miscellaneous
metal and plastic workers; team assemblers; slaughterers
and meat packers; electrical and electronic equipment assemblers; and production first-line supervisors.
Employment in computer and mathematical science
occupations shrank after layoffs for at least four of the
layoff reasons related to business demand, and business
and financial occupations and architecture and engineering occupations lost employment from layoffs due to at
least four of the reasons.
Layoffs because of slack work resulted in employment
declines in many occupational groups, about a third of
which were production jobs. Most affected were metal
and plastic workers, team assemblers, production supervisors, electrical equipment assemblers, and sewing machine
operators. The occupations that grew were hand laborers, computer applications software engineers, customer
service representatives, stock clerks, and market research
analysts.
After layoffs due to excess inventory and domestic competition, overall employment levels shrank in every occupational group in which employment had been reported
before the layoffs; cutbacks occurred in both core and
noncore occupations regardless of business function.
Financial. Financial-related reasons for layoffs include
financial difficulty; bankruptcy; and measures to control
costs, cut costs, and increase profitability. As with layoffs
related to business demand, financial-related layoffs resulted in job losses among skilled workers, in addition to
losses among less skilled workers. Notable outcomes were
a large decline in personal care and service occupations
after bankruptcy, sales workers after cost control layoffs,
and architecture and engineering occupations after financial difficulty.
The largest employment declines in layoffs due to general financial difficulty were in production, transportation and material moving, and office and administrative
support occupations. Production occupations with job
losses included inspectors, testers, and weighers; team assemblers; and production supervisors. This type of layoff
also resulted in the employment of fewer transportation
workers; recordkeeping weighers, measurers, checkers,
and samplers; flight attendants; and parking lot attendants. Among computer and engineering jobs lost were
computer programmers and computer systems software
engineers, and applications engineers. The same set of
establishments eventually hired workers for computer
science occupations that were less research intensive in
nature: computer systems analysts; network support and
data communications analysts; and network and computer systems administrators. They also hired many more
registered nurses, cashiers, and accountants and auditors.
Occupations that experienced employment cutbacks after bankruptcy were reservation and transportation ticket
agents, stock clerks and order fillers, industrial truck and
tractor operators, vehicle and equipment cleaners, and
electronics engineers (except computer), among others.
Cost control and cost cutting resulted in large employment
declines, in terms of both percentages and levels, among
architecture and engineering occupations, but the change
was concentrated in a few establishments. Establishments
with this type of layoff had employment declines in several administrative support and sales occupations directly
related to sales functions: customer service representatives; shipping, receiving, and traffic clerks; stock clerks
and order fillers; and retail salespersons. These establishments also cut some production and maintenance workers
who tended to be paid higher wages: supervisors of mechanics, installers, and repairers; transportation managers;
and supervisors of production workers.
Some administrative support occupations whose employment grew after cost-cutting layoffs were those related
to internal staffing and support: payroll and timekeeping
clerks, human resources workers, administrative support
supervisors, general office clerks, and bookkeeping and accounting clerks. In addition, employers whose layoffs were
a result of controlling or cutting costs hired workers for
several laborer occupations that tended to be paid lower
wages: hand laborers and freight and stock movers; and
janitors and cleaners.
Production. The kinds of jobs lost from production-related extended mass layoffs related to the specific reason
cited for the layoff. Although production worker employment shrank overall, it grew in establishments whose layoffs had been due to automation or technological advances. Transportation and material moving occupations grew
in establishments whose layoffs had been due to model
changeover or product line discontinuations.
After product line discontinuation, employment changes
in a few large establishments accounted for the large decreases in production worker employment. Occupations
that shrank included slaughterers and meat packers, assemblers and fabricators, transportation equipment painters, synthetic and glass fiber machine setters, inspectors
and weighers, welders, semiconductor processors, and
engine assemblers. Production occupations that grew included production worker helpers, packaging and filling
machine operators, bakers, coating and painting machine
operators, and upholsterers.
Layoffs due to plant or machine repair or maintenance
tended to affect occupations directly related to the operation of machines and production systems, and more
production workers were eventually added than dropped.
Occupations whose employment decreased included inspectors and weighers, extruding and compacting machine
operators, furnace operators and tenders, and chemical
plant and system operators. Occupations with employment increases included metal and plastic drilling and
boring machine tool operators; meat, poultry, and fish cutters and trimmers; cleaning and metal pickling equipment
operators; and coating and spraying machine operators.
Occupations whose employment declined after automation/technological advances provide insight into the types
of jobs at risk as technology advances. Declines occurred
among engine machine assemblers, machine feeders and
offbearers, metal and plastic computer-controlled machine tool operators, tool and die makers, data entry keyers, and tool grinders.
Disaster/safety. Disaster/safety concerns comprised a
hazardous work environment, a natural (not weather related) disaster, a nonnatural disaster, and extreme weather-related events. Extreme weather-related events were
responsible for most of the employment declines in this
category. Jobs lost after layoffs that were due to extreme
Monthly Labor Review • June 2011 23
Mass Layoffs and Employment
weather-related events affected primarily service workers
providing transportation (transit and intercity bus drivers), security (security guards), food service (waiters and
waitresses and restaurant cooks), housekeeping (maids,
housekeeping cleaners, and janitors), and entertainment
(gaming dealers, and tour guides and escorts); many of
these occupations may be affected by tourism. Job gains
were in construction occupations.
Domestic and overseas relocation. Although the sample
size of establishments that laid off workers due to domestic and overseas relocation is smaller than the sample for
other layoff reasons (MLS ended the two series with the
2003 data), the study sample still had almost 300 units reporting under the former reason and more than 200 units
under the latter—enough to study the outcomes of layoffs
for these reasons.
After layoffs due to overseas relocation, only two occupational groups grew in employment: office and administrative support; and arts, design, entertainment, sports, and
media. Detailed occupations that shrank included various
assemblers, machine operators, hand laborers, and industrial and electronic engineers and their managers. Despite
the reductions among major occupational groups, the establishments hired workers in occupations related to sales,
shipping, human resources, and computer network support (such as stock, billing, and shipping clerks; sales representatives; network administrators; and human resource
specialists).
After layoffs due to domestic relocation,21 establishments
reduced employment in two of the higher skilled groups:
healthcare practitioner and technical occupations; and life,
physical, and social science occupations. Establishments
reduced employment in occupations involving nonanalytical skill, such as production; office and administrative
support; transportation and material moving; installation,
maintenance, and repair; protective service; building and
grounds cleaning and maintenance; construction and extraction; and sales and related occupations. In contrast,
the establishments with layoffs due to domestic relocations hired more analytical occupations: business and
financial operations; architecture and engineering; management; arts, design, entertainment, sports, and media;
computer and mathematical science; as well as personal
care and service occupations.
Occupational changes by geographic region
Without regression analysis. The effect of layoffs on occupational employment levels varied across the country
24 Monthly Labor Review • June 2011
because of differences in industry composition, local labor market conditions, and other economic factors. In the
entire MLS data set (the universe of mass layoff events
during the 2000–2007 period), the Midwest, West, and
Southeast regions had the most worker separations, with
losses of 2.5, 2.2, and 1.2 million jobs, respectively. The
New York-New Jersey, Mid-Atlantic, and Southwest regions had between 500,000 and 900,000 separations. New
England and the Mountain-Plains region had the fewest number of worker separations, each less than 500,000
over the same period.
Table 12 shows the percent change in employment after
layoffs, by geographic region and occupational group. The
States within each region are shown in figure 1.
Occupational groups that shrank in employment across
most regions involved administration, personal service,
and mainly nonanalytical skills: production, sales, office
and administrative support, protective service, management, transportation, installation, construction, personal
care, and building maintenance. Employment in occupation groups involving analytical skills grew in more regions than did other occupations. Business and financial
operations, legal, computer and mathematical science,
healthcare practitioner and technical, and community and
social service occupations had lower percent declines in
employment in most regions relative to declines among
other occupational groups.
Within regions, the pattern of reducing employment
in occupations involving clerical and nonanalytical skills
while retaining jobs requiring analytical skills was most
prevalent in the Mid-Atlantic, New England, New YorkNew Jersey, and Southwest regions. The first three regions
have high concentrations of industries that experienced
relatively large numbers of layoffs. Establishments undergoing layoffs in the Mid-Atlantic region shed team assemblers; data entry keyers; shipping, receiving, and order
clerks; retail salespersons; and hand laborers. The MidAtlantic establishments added financial analysts, industrial machinery mechanics, and computer systems analysts.
The primary finding in this regional analysis was that
when entire industries retained or increased employment in occupations core to their business, the pattern
also manifested itself throughout most of the geographic
regions. In 6 of the 8 geographic areas—all except the
Southeast and Southwest—employers either added workers or lost fewer workers in job functions core to the industries dominant in their economies. The industry distribution of each region’s employment was used to estimate
which industries were dominant.22
In the Midwest region, which had the most worker
Table 12. Percent change in employment after extended mass layoff, by geographic region and occupational group, 1999–2008
Occupational group
New
England
New
York–New
Jersey
Mid–
Atlantic
Southeast
Midwest
Southwest
Mountain–
Plains
West
Management
Business and financial operations
Computer and mathematical science
Architecture and engineering
Life, physical, and social science
Community and social services
Legal
Education, training, and library
Arts, design, entertainment, sports, and media
Healthcare practitioners and technical
Healthcare support
Protective service
Food preparation and serving related
Building and grounds cleaning and maintenance
Personal care and service
Sales and related
Office and administrative support
Construction and extraction
Installation, maintenance, and repair
Production
Transportation and material moving
–38.5
70.4
–25.6
–63.8
–69.5
.0
19.57
–35.6
–78.1
(1)
(1)
(1)
(1)
–16.8
(1)
–64.0
–29.7
3.1
–39.1
–52.8
–3.0
24.7
20.1
35.9
–16.0
17.8
3.7
49.8
21.0
26.7
5.2
26.2
–40.4
.7
–7.6
–20.3
–38.0
–21.9
–13.0
–11.7
–17.6
–29.8
–47.2
–21.5
31.7
–26.4
261.4
67.2
.9
–30.8
27.7
–.8
(1)
–12.1
–42.6
–30.8
–39.5
–23.9
–30.5
–16.7
–28.5
–28.1
–26.9
–26.2
–12.6
–.5
–9.0
–7.3
16.8
7.5
–33.3
–58.0
–3.4
12.6
–32.6
67.0
24.1
–30.9
–24.8
–21.4
–2.6
–25.8
–31.3
–28.8
–18.0
–7.4
–5.5
–20.9
20.6
–2.7
56.4
4.4
1.9
–9.9
–24.0
–8.0
–27.9
–3.9
–3.7
–11.9
–9.1
–14.0
–14.4
–18.3
–12.3
–34.2
8.4
–13.3
–4.9
–50.4
–6.8
–27.8
4.1
–22.0
79.6
78.5
–20.5
–1.3
–.8
–3.1
–22.6
–16.2
–11.7
27.2
–4.9
–14.6
–39.9
35.7
–3.0
–9.1
–44.1
–10.6
44.1
–2.2
–12.4
–13.5
–34.4
–63.0
–20.0
–50.4
–9.6
–19.7
–21.8
–15.1
–15.9
–24.3
9.6
–15.5
–3.3
–16.4
–17.5
–31.1
1.8
–2.5
8.5
–8.7
6.3
–11.3
–6.0
1.0
–2.6
–4.2
–13.7
–23.8
–.1
–21.4
–19.8
–16.7
1
Percent change excluded because it is based on fewer than 5 establishments reporting occupations in the occupational group before layoffs
separations, food preparation and serving-related occupations composed the group that experienced the largest percent decline in employment. The Midwest division23 had relatively high employment concentrations in
manufacturing (15.7 percent) and wholesale and retail
trade (15.2 percent) in 2004. After layoffs, Midwest region employment in the occupations core to these sectors—production, sales, and transportation and material
moving occupations—showed relatively small losses. The
other regions had larger percent declines in employment
in these occupations.
Likewise, the Pacific division24 had relatively high concentrations of employment in the information sector (3.0
percent)—particularly in motion picture and sound recording industries—and the study’s West region lost relatively little employment in the arts, design, entertainment,
and media occupations, which are core to this industry.
New York and New Jersey, which both have relatively
high proportions of their employment in the information and financial activities sectors, tended to hire workers
in the sectors’ core occupations.25 Employers in the New
York-New Jersey region hired workers in analytical occupations such as computer applications engineers, management and computer systems analysts, accountants,
and industrial engineers; they let go of retail salespersons,
parking lot attendants, team assemblers, hand laborers,
and telemarketers.
Employers in New England cut jobs of office and production worker supervisors, general maintenance workers,
janitors and cleaners, and stock clerks. Meanwhile, they
hired accountants and financial managers. New England
had relatively high concentrations of employment in the
education (9.8 percent), health care (14.4 percent), and
finance and insurance (6 percent) sectors. After layoffs,
businesses in the region added workers in occupations
that are core to some of these industries; employment in
business and financial operations occupations increased
by 70 percent in the study group.
Employers in the Mountain-Plains region also followed
the pattern of shedding relatively few workers in occupations core to the industries that dominate the region’s
economy. In 2004, among all Census regions, the Mountain division26 was the geographic area with the highest
concentrations of employment in leisure and hospitality
(11.2 percent). Perhaps, as a result, the region had lower
percent declines in employment in food preparation, personal care, and sales occupations than did other regions.
As noted earlier, however, not every region followed the
pattern of retaining or increasing jobs in occupations core
to the region’s dominant industries, in part due to changing technology, consumer trends, or business practices.
The combined South Atlantic and East South Central
Census divisions—together approximating the study’s
Southeast region27—had relatively high employment conMonthly Labor Review • June 2011 25
Mass Layoffs and Employment
Figure 1. Definition of regions by state, as used in this study
Mountain-Plains—includes Colorado, Kansas,
Missouri, Montana, Utah,
and Wyoming
Midwest—includes Illinois,
Indiana, Iowa, Michigan,
Minnesota, Nebraska, North
Dakota, Ohio, South Dakota,
and Wisconsin
New England—includes
Connecticut, Maine, Massachusetts, New Hampshire, Rhode
Island, and Vermont
New YorkNew Jersey—
includes New
York and New
Jersey
West—
includes
Alaska,
Arizona,
California,
Hawaii, Idaho,
Nevada,
Oregon, and
Washington
Mid-Atlantic—includes
Pennsylvania, Delaware,
District of Columbia,
Maryland, Virginia, and
West Virginia
Southeast—includes Alabama,
Florida, Georgia, Kentucky, Mississippi, North Carolina, South
Carolina, and Tennessee
Southwest—includes
Arkansas, Louisiana, New
Mexico, Oklahoma, and Texas
centrations in the transportation and utilities industries
but lost a comparatively large number of jobs in transportation and material moving occupations. Similarly, the
Southwest region deviated from the pattern of regions retaining occupations core to dominant industries. Relative
to other regions, the Census-defined West South Central
division28 had a high proportion of employment in the
mining sector (1.5 percent) and the telecommunications
industry (1.2 percent). In May 2008, telecommunications
businesses were one of the largest employers of telemarketers. After layoffs, however, the study’s Southwest region had fewer telemarketers.
Regression analysis. A final regression analysis was conducted to analyze the effect of layoffs on occupational employment by geographic region, controlling for industry,
time between observations, and establishment size. The
variables of interest were the eight interaction terms for
layoff × region. The model included non-interaction dummies for all regions except the Mountain-Plains region, so
the interaction terms did not combine the impacts of being located in a particular region with having layoffs in
that region. The regression was based on a total of 209,858
observations: 205,339 control observations and 4,520
study observations.
26 Monthly Labor Review • June 2011
The model used was
∆employmentSOC major group = β0 + β1 layoffi × West +
β2 layoffi × Southwest + β3 layoffi × Southeast + β4 layoffi
× Mountain-Plains + β5 layoffi × New York-New Jersey
+ β6 layoffi × Midwest + β7 layoffi × Mid-Atlantic + β8
layoffi × New England + β9goods + β10totalemp_firsti +
ΣjδjI(geographic regionji) + ΣjγjI(number of years between
observationsji) + εi,
where the dependent variable was the change in employment level for each occupational group between the first
and second observations. Layoff was a dummy variable
indicating whether the establishment had a layoff; where
layoffi = 1, a layoff occurred. The geographic region dummy
variables indicated geographic region: West, Southwest,
Southeast, Mountain-Plains, New York-New Jersey,
Midwest, Mid-Atlantic, and New England. Geographic
region included non-interaction dummies for all regions
except the Mountain-Plains region. Layoff × [region] was
an interaction dummy variable. Goods was a dummy variable for the goods-producing aggregation, as opposed to
the service-providing aggregation. Totalemp_first was the
total employment in the establishment at the time of the
first observation. Finally, there were nine dummy variables
representing the number of years between observations,
ranging from 1 to 9; 9 years was captured in the intercept.
In general, the Mid-Atlantic region saw the most
substantial employment change for many occupational
groups, compared with other regions. Appendix table A-4
shows the regression output for the eight regional interaction variables. For seven occupational groups, an extended
mass layoff was associated with a decline in additional
employment in all geographic areas (where a change was
statistically significant). Production employment shrank
in every region, most noticeably in the Mid-Atlantic, followed by the Southwest, Mountain-Plains, New England,
and Midwest regions. Production employment declined
the least in New York-New Jersey. Employment in transportation and material moving occupations declined the
most in the Mid-Atlantic, Southeast, and New York-New
Jersey regions. It grew only in the Mountain-Plains. Employment in office and administrative support occupations declined the most in the Mid-Atlantic region and
the least in the Mountain-Plains.
Similarly, employment in sales and related occupations
declined in all regions; the largest decline was in New
York-New Jersey and the Southeast, and the smallest was
in the Midwest. Construction and extraction employment
shrank everywhere (where a change was statistically significant), especially in the Midwest. Finally, architecture
and engineering employment declined substantially in
New England and fell the least in the Southeast.
For four occupational groups, an extended mass layoff
was associated with a decline in additional employment in
nearly all geographic areas (where statistically significant).
Healthcare practitioners and technical employment grew
only in the Southwest and actually declined in the Midwest and Southeast. Similarly, healthcare support employment grew only in the Southwest and declined in the
West and Midwest. Management employment declined
the most in the Mid-Atlantic and Mountain-Plains, and
grew only in New York-New Jersey. Employment in the
arts, design, entertainment, sports, and media occupational group declined the most in the Southeast and New
England, and grew only in New York-New Jersey.
In contrast, for three occupational groups, an extended
mass layoff was associated with an increase in additional
employment in all geographic areas (where statistically
significant). Employment in legal occupations grew the
most in New York-New Jersey. The Southeast was the
region with the most growth in building and grounds
cleaning and maintenance employment, and also in food
preparation and serving occupations employment.
Finally, the regression yielded mixed results for five occupational groups. Business and financial operations grew
the most in New England and the Mountain-Plains, but
declined in the Southeast and Midwest. Computer and
mathematical science employment grew the most in the
Mid-Atlantic and New York-New Jersey regions, but
declined in the West and Southwest. Life, physical, and
social science employment grew the most in the Mid-Atlantic, and declined the most in the Southwest. Personal
care and service employment grew in the Midwest and
declined the most in the Southeast. Installation, maintenance and repair employment grew in the Southwest and
shrank the most in the Southeast and New England. Finally, protective service employment grew in the Midwest
but declined in New York-New Jersey and the Southeast.
DURING THE PERIOD COVERED BY THIS STUDY, the
economy experienced both the dot-com bubble burst and
large numbers of layoffs in manufacturing. The occupations that were most affected in establishments with layoffs were those which generally involved nonanalytical
skills and abilities and tended to represent business support functions, while the occupations whose employment
level was relatively unaffected by the layoffs were those
which involved analytical skills and abilities or were core
to their business. Today’s labor market turbulence can be
found in different industries, such as real estate and finance. Repeating the study with newer data would reveal
whether the patterns observed in this analysis hold under
different economic circumstances. In addition, using data
further from the layoff date, rather than the first observation after the first layoff, might reveal different long-term
restructuring outcomes.
NOTES
ACKNOWLEDGEMENTS: The authors thank the following BLS
employees for their guidance: Patrick Carey, Elizabeth Weber Handwerker, Michael Soloy, Dixie Sommers, George Stamas, Laura Train,
and Zachary Warren.
1
According to the MLS, during the first quarter of 2009, there
were 3,979 extended mass layoff events, resulting in the separation of
705,141 workers from their jobs for at least 31 days. In the first quarter
of 2011, there were 1,397 mass layoff events that resulted in 190,895
separations. Extended mass layoff events and separations have shown
an over-the-year decrease for six consecutive quarters. BLS Mass Layoff Statistics are available at http://www.bls.gov/news.release/mslo.
toc.htm (visited June 28, 2011). Extended mass layoff data have been
available since second quarter 1995.
Monthly Labor Review • June 2011 27
Mass Layoffs and Employment
Recessions are identified by the National Bureau of Economic Research (NBER). For a list of recession start and end dates, see “U.S.
Business Cycle Expansions and Contractions” (Cambridge, MA, National Bureau of Economic Research, June 20, 2011), http:/www.nber.
org/cycles/cyclesmain.html (visited June 20, 2011).
2
3
This article uses the term “function” differently than does the MLS
program, so the data are not comparable.
4
Approximately 30 days after a mass layoff begins, the employer is
contacted for additional information.
5
The universe of OES establishments from which the study sample
was drawn includes only usable units that passed all BLS tests and that
reported all requested employment data and all or partial wage data.
Farming, fishing, and forestry occupations were included in the total calculations, but were not included in the analysis, because MLS and
OES data include only nonfarm industries.
6
7
See “O*NET OnLine: Browse by O*NET Data,” http://online.
onetcenter.org/find/descriptor/browse/Abilities (visited July 26,
2010). O*NET identifies six descriptors (categories of occupational information): knowledge, skills, abilities, work activities, interests, and
work values. Each descriptor has a set of elements, and each element
has importance scores for all O*NET occupations.
Occupations in the “analytical” group have high importance scores in
skills such as reading comprehension; writing; speaking; math; science;
critical thinking; complex problem solving; judgment and decision
making; systems analysis; active listening; monitoring; social perceptiveness; coordination; persuasion; negotiation; instructing; service orientation; and management of time, financial, material, and personnel
resources. “Analytical” occupations also have high importance scores in
elements such as deductive and inductive reasoning, fluency of ideas,
informative ordering, mathematical reasoning, memorization, number
facility, oral and written expression and comprehension, perceptual
speed, and problem sensitivity.
Occupations in the “nonanalytical” group have high importance
scores in nonanalytical skills such as equipment maintenance, troubleshooting, repairing, and quality control analysis. These occupations
have high importance scores in nonanalytical abilities such as dynamic
and extent flexibility; dynamic, explosive, static, and trunk strength;
gross body coordination and equilibrium, and stamina. “Nonanalytical” occupations also tend to possess elements of psychomotor abilities,
such as control precision and manual dexterity.
Education and training data come from “Employment Projections:
EPP Tables—Occupations”(U.S. Bureau of Labor Statistics, no date),
table 1.11, http://www.bls.gov/emp/#tables, (visited July 27, 2010).
This observation references an occupation’s distribution of employment by educational attainment (found in the table). See also Occupational Outlook Handbook, 2010–11 Edition (U.S. Bureau of Labor
Statistics, no date), http://www.bls.gov/oco (visited June 13, 2011).
Information about general office clerks, customer service representatives, and secretaries and administrative assistants is also from the
Handbook, at http://www.bls.gov/oco/ocos130.htm#training, http://
www.bls.gov/oco/ocos280.htm#training, and http://www.bls.gov/
oco/ocos151.htm#training, respectively (visited July 27, 2010).
8
9
“Employment Projections: EPP Tables—Occupations.”
10
See “Assemblers and Fabricators,” in Occupational Outlook Handbook, 2010–11 Edition, http://www.bls.gov/oco/ocos217.htm (visited
July 27, 2010).
See Career Guide to Industries, 2010–11 Edition (Bureau of Labor Statistics, no date), http://www.bls.gov/oco/cg (visited June 13,
11
28 Monthly Labor Review • June 2011
2011).
12
See “Truck Drivers and Driver/Sales Workers: Training, Other
qualifications, and Advancement,” in Occupational Outlook Handbook,
2010–11 Edition, http://www.bls.gov/oco/ocos246.htm#training
(visited July 27, 2010).
13
“Employment Projections: EPP Tables—Occupations.”
The reduced employment in management occupations overall may
be, in part, a result of improvements in the classification of managers
in the OES survey. The interpretation of the employment change in
management occupations should be made with caution.
14
Layoffs because of contract completion or contract cancellation
could be attributable to seasonal factors, but, for the purposes of this
study, they were included in the “economic difficulties” category.
15
16
Because the OES program surveys the same establishment at the
same time each year, staffing pattern changes in the seasonal layoffs category are long-term changes rather than the result of seasonal changes.
See “Mass Layoff Statistics” (U.S. Bureau of Labor Statistics, no
date), http://www.bls.gov/mls (visited Mar. 18, 2010).
17
MLS ended the domestic/foreign relocation series with the 2003
data. Relocation of work was then replaced by movement of work data.
The category of controlling costs was added as a reason in 2007.
18
19
The rankings excluded seasonal layoffs—although they were included in the count of total mass layoffs.
20
The rankings excluded seasonal layoffs, refusal to respond, “other,”
and “does not know” as reasons for layoffs—although they were included in the count of total mass layoffs.
The data relating to domestic relocation reflect occupations with at
least 10 establishments reporting them initially.
21
From Geographic Profile of Employment and Unemployment,
2004, “Table 7. Census regions and divisions: percent distribution of
employed persons by industry, sex, race, and Hispanic or Latino ethnicity, 2004 annual averages” (U.S. Bureau of Labor Statistics, January 2009), also available online at http://www.bls.gov/opub/gp/pdf/
gp04_07.pdf (visited June 13, 2011).
22
The Midwest division is similar to the study’s Midwest region but
also includes Missouri and Kansas.
23
The Pacific division is similar to the study’s West region but excludes Arizona, Idaho, and Nevada.
24
25
From Geographic Profile of Employment and Unemployment, 2004,
“Table 20. States: percent distribution of employed persons by sex,
race, Hispanic or Latino ethnicity, and industry, 2004 annual averages”
(U.S. Bureau of Labor Statistics, January 2009), also available online
at http://www.bls.gov/opub/gp/pdf/gp04_20.pdf (visited June 13,
2011).
26
The Mountain division is similar to the study’s Mountain-Plains
region but excludes Missouri and Kansas and includes Arizona and
New Mexico.
27
The South Atlantic and East South Central Census divisions are
similar to the study’s Southeast region, except that the combined Census-defined divisions include Virginia, Delaware, the District of Columbia and West Virginia; these four jurisdictions are all in the study’s
Mid-Atlantic region.
28
The West South Central division is similar to the study’s Southwest region but excludes New Mexico.
Appendix tables: Output for regressions
Table A-1. Output for 21 regressions of occupational group on layoffi and control variables (industry, region, establishment
size, and number of years between observations) Dependent variable (change in employment in occupational
group)
Production
Office and administrative support
Sales and related
Management
Transportation and material moving
Architecture and engineering
Installation, maintenance, and repair
Construction and extraction
Education
Personal care and service
Arts, design, entertainment, sports, and media
Computer and mathematical science
Healthcare practitioner and technical
Protective service
Healthcare support
Life, physical, and social science
Legal
Community and social services
Business and financial operations
Building and grounds cleaning and maintenance
Food preparation and serving related
Layoffi
parameter
estimate
Standard error
on layoffi
parameter
estimate
t-value
p-value
–20.76
–10.97
–7.66
–6.13
–6.04
–4.29
–3.38
–1.92
–1.61
–1.28
–1.13
–.48
–.05
–.04
–.03
–.03
.60
.68
1.28
2.62
5.36
1.00
1.23
.51
.43
.79
.54
.46
.43
1.54
.61
.32
.67
.78
.40
.39
.34
.11
.44
.48
.37
.58
–20.86
–8.91
–15.05
–14.27
–7.69
–7.96
–7.41
–4.42
–1.05
–2.1
–3.51
–.71
–.06
–.09
–.08
–.09
5.32
1.54
2.66
7.09
9.27
<.0001
<.0001
<.0001
<.0001
<.0001
<.0001
<.0001
<.0001
.2938
.0356
.0004
.4792
.9513
.9263
.9371
.9309
<.0001
.1225
.0078
<.0001
<.0001
NOTE: Table excludes output for control variables to conserve space, and excludes farming, fishing, and forestry occupations. Degrees of freedom=1
for all regressions.
Monthly Labor Review • June 2011 29
Mass Layoffs and Employment
Table A-2. Output for 21 regressions of occupational group on economici, seasonali, and control variables (industry,
region, establishment size, and number of years between observations)
Dependent variable
(change in employment in
occupational group)
Management
Business and financial
operations
Computer and
mathematical science
Architecture and
engineering
Life, physical, and social
science
Community and social
services
Legal
Education
Arts, design, entertainment,
sports, and media
Healthcare practitioner and
technical
Healthcare support
Protective service
Food preparation and
serving related
Building and grounds cleaning and maintenance
Personal care and service
Sales and related
Office and administrative
support
Construction and extraction
Installation, maintenance,
and repair
Production
Transportation and material
moving
Interaction term
layoffi*economici
layoffi*seasonali
layoffi*economici
layoffi*seasonali
layoffi*economici
layoffi*seasonali
layoffi*economici
layoffi*seasonali
layoffi*economici
layoffi*seasonali
layoffi*economici
layoffi*seasonali
layoffi*economici
layoffi*seasonali
layoffi*economici
layoffi*seasonali
layoffi*economici
layoffi*seasonali
layoffi*economici
layoffi*seasonali
layoffi*economici
layoffi*seasonali
layoffi*economici
layoffi*seasonali
layoffi*economici
layoffi*seasonali
layoffi*economici
layoffi*seasonali
layoffi*economici
layoffi*seasonali
layoffi*economici
layoffi*seasonali
layoffi*economici
layoffi*seasonali
layoffi*economici
layoffi*seasonali
layoffi*economici
layoffi*seasonali
layoffi*economici
layoffi*seasonali
layoffi*economici
layoffi*seasonali
Coefficient
Standard
error
t-value
p-value
0.00
–.65
3.73
.12
1.36
–1.56
.69
3.53
–.32
.68
–.33
–.14
.04
.11
–1.31
.88
6.11
–3.47
7.60
6.62
3.03
2.45
3.41
–1.05
–9.22
6.21
–.18
4.36
7.66
–.60
–4.80
–4.70
7.87
11.47
–4.50
–6.86
3.26
4.52
–15.62
.10
–2.67
–4.55
3.49
3.68
4.11
4.33
5.43
5.72
5.80
6.12
2.05
2.16
.69
.73
.38
.40
2.13
2.25
4.15
4.37
6.93
7.31
1.85
1.94
1.74
1.84
7.75
8.16
2.80
2.95
5.65
5.96
5.90
6.22
9.58
10.10
4.07
4.29
5.28
5.57
12.30
12.96
7.47
7.88
0
–.18
.91
.03
.25
–.27
.12
.58
–.16
.31
–.48
–.19
.1
.27
–.61
.39
1.47
–.79
1.1
.91
1.64
1.26
1.96
–.57
–1.19
.76
–.07
1.48
1.36
–.1
–.81
–.76
.82
1.14
–1.11
–1.6
.62
.81
–1.27
.01
–.36
–.58
0.9996
.8608
.3633
.9781
.8022
.7858
.906
.5645
.8767
.7536
.6341
.8514
.9174
.7853
.539
.6963
.1408
.4279
.2726
.365
.1007
.2078
.0505
.5689
.2341
.4465
.9476
.1393
.1752
.9197
.4167
.4499
.4114
.2563
.2686
.1098
.5374
.4169
.2041
.9941
.721
.5631
Are the two
interaction terms
significantly
different?
no
no
no
no
no
no
no
no
no
no
no
no
no
no
no
no
no
no
no
no
no
NOTE: Table excludes output for control variables to conserve space, and excludes farming, fishing, and forestry occupations. Degrees of
freedom=1 for all regressions.
30 Monthly Labor Review • June 2011
Table A-3. Output for 21 regressions of occupational group on layoffi goods–producing group, layoffi service­–providing
group, and control variables (region, establishment size, years between observations, and goods-producing
group)
Dependent variable
(change in employment in
occupational group)
Management
Business and financial
operations
Computer and mathematical
science
Architecture and
engineering
Life, physical, and social
science
Community and social
services
Legal
Education
Arts, design, entertainment,
sports, and media
Healthcare practitioner and
technical
Healthcare support
Protective service
Food preparation and
serving related
Building and grounds
cleaning and maintenance
Personal care and service
Sales and related
Office and administrative
support
Construction and extraction
Installation, maintenance,
and repair
Production
Transportation and material
moving
Interaction term
layoffi* goodsi
layoffi* servicei
layoffi* goodsi
layoffi* servicei
layoffi* goodsi
layoffi* servicei
layoffi* goodsi
layoffi* servicei
layoffi* goodsi
layoffi*servicei
layoffi* goodsi
layoffi* servicei
layoffi* goodsi
layoffi* servicei
layoffi* goodsi
layoffi* servicei
layoffi* goodsi
layoffi* servicei
layoffi* goodsi
layoffi* servicei
layoffi* goodsi
layoffi* servicei
layoffi* goodsi
layoffi* servicei
layoffi* goodsi
layoffi* servicei
layoffi* goodsi
layoffi* servicei
layoffi*goodsi
layoffi* servicei
layoffi* goodsi
layoffi* servicei
layoffi* goodsi
layoffi* servicei
layoffi* goodsi
layoffi* servicei
layoffi* goodsi
layoffi* servicei
layoffi* goodsi
layoffi* servicei
layoffi*goodsi
layoffi*servicei
Coefficient
Standard
error
t–value
p–value
–4.94
–7.12
.85
1.65
–.92
–.11
–7.45
–1.66
–.51
.37
.82
.57
.62
.58
–1.13
–2.01
.01
–2.06
–1.29
.98
–.11
.03
1.12
–1.00
4.47
6.09
2.17
2.99
1.82
–3.84
–1.87
–12.44
–1.28
–18.99
–4.33
.08
–5.16
–1.92
–48.17
1.91
–5.83
–6.22
0.55
.52
.62
.58
.86
.81
.69
.65
.44
.41
.57
.53
.14
.14
1.98
1.85
.41
.39
1.01
.94
.50
.46
.52
.48
.74
.70
.47
.44
.78
.73
.65
.61
1.58
1.48
.56
.52
.59
.55
1.28
1.19
1.01
.95
–8.92
–13.77
1.36
2.83
–1.06
–.13
–10.76
–2.57
–1.16
.9
1.44
1.07
4.27
4.31
–.57
–1.09
.01
–5.34
–1.27
1.03
–.22
.07
2.17
–2.06
6.01
8.75
4.57
6.72
2.32
–5.24
–2.86
–20.32
–.81
–12.82
–7.76
.15
–8.77
–3.48
–37.71
1.6
–5.77
–6.58
<.0001
<.0001
.173
.0046
.2869
.8948
<.0001
.0102
.2445
.3694
.1503
.2838
<.0001
<.0001
.5685
.2761
.9898
<.0001
.203
.3015
.8277
.942
.0298
.0392
<.0001
<.0001
<.0001
<.0001
.0204
<.0001
.0042
<.0001
.4202
<.0001
<.0001
.8812
<.0001
.0005
<.0001
.1093
<.0001
<.0001
Are the parameter
estimates significantly different at
the 5-percent significance level?
yes
no
no
yes
yes
no
no
no
yes
no
no
yes
no
yes
yes
yes
yes
yes
yes
yes
no
NOTE: Table excludes output for control variables to conserve space, and excludes farming, fishing, and forestry occupations. Degrees of freedom=1 for all regressions.
Monthly Labor Review • June 2011 31
Mass Layoffs and Employment
Table A–4. Output for regressions of change in employment in each occupational group on all 8 layoff region interaction
variables, and controls
Dependent variable
(change in employment in occupational group)
Output for
interaction
terms
Interaction term
New
England
New York/
New Jersey
Mid–
Atlantic
Southeast
Midwest
Southwest
Mountain–
Plains
West
Management
Coefficient
Standard error
p–value
–12.26
3.85
.00
3.81
1.06
.00
–14.92
2.23
<.0001
–7.06
.84
<.0001
–5.55
.67
<.0001
–10.03
1.04
<.0001
–13.47
1.35
<.0001
–5.80
.67
<.0001
Business and
financial operations
Coefficient
Standard error
p–value
20.99
4.32
<.0001
6.32
1.19
<.0001
–4.36
2.50
.08
–2.06
.95
.03
–.92
.76
.23
2.87
1.17
.01
11.57
1.52
<.0001
.16
.75
.83
Computer and
mathematical
science
Coefficient
Standard error
p–value
–8.69
6.02
.15
9.13
1.66
<.0001
9.78
3.49
.01
.50
1.32
.70
–.15
1.05
.88
–3.82
1.62
.02
–1.00
2.12
.64
–3.98
1.04
.00
Architecture and
engineering
Coefficient
Standard error
p–value
–11.21
4.82
.02
–2.81
1.33
.03
–2.65
2.79
.34
–2.05
1.06
.05
–5.25
.84
<.0001
–1.85
1.30
.16
–3.89
1.70
.02
–6.03
.84
<.0001
Life, physical, and
social science
Coefficient
Standard error
p–value
–1.15
3.05
.71
1.49
.84
.08
8.60
1.77
<.0001
–.05
.67
.94
1.36
.53
.01
–2.04
.82
.01
–1.54
1.07
.15
–1.40
.53
.01
Community and
social service
Coefficient
Standard error
p–value
.52
3.97
.90
.23
1.09
.83
1.74
2.30
.45
1.30
.87
.13
.74
.70
.29
.49
1.07
.64
.71
1.40
.61
.45
.69
.51
Legal
Coefficient
Standard error
p–value
.61
1.01
.54
1.01
.28
.00
.54
.58
.35
.56
.22
.01
.79
.18
<.0001
.22
.27
.42
.75
.36
.03
.39
.17
.03
Education
Coefficient
Standard error
p–value
–2.69
13.76
.85
3.00
3.79
.43
–3.56
7.97
.66
–4.65
3.01
.12
–.61
2.41
.80
–3.11
3.71
.40
–3.36
4.84
.49
–1.48
2.38
.54
Arts, design,
entertainment,
sports, media
Coefficient
Standard error
p–value
–5.61
2.87
.05
1.45
.79
.07
2.14
1.67
.20
–6.31
.63
<.0001
.18
.50
.72
–1.03
.78
.19
–.56
1.01
.58
–.81
.50
.10
Healthcare
practitioners and
technical
Coefficient
Standard error
p–value
–1.11
7.02
.87
–1.07
1.94
.58
–2.99
4.07
.46
–2.78
1.54
.07
–3.12
1.23
.01
17.20
1.90
<.0001
–2.61
2.47
.29
–.68
1.22
.58
Healthcare support
Coefficient
Standard error
p–value
–.13
3.45
.97
.79
.95
.40
.27
2.00
.89
–.11
.75
.89
–1.19
.60
.05
4.74
.93
<.0001
–.20
1.21
.87
–.90
.60
.13
Protective service
Coefficient
Standard error
p–value
.38
3.60
.92
–1.55
.99
.12
1.29
2.09
.54
–1.63
.79
.04
1.30
.63
.04
–.02
.97
.99
–1.61
1.27
.20
.34
.62
.58
Food preparation
and serving related
Coefficient
Standard error
p–value
3.53
5.18
.50
4.56
1.43
.00
–2.83
3.00
.35
15.58
1.13
<.0001
2.29
.91
.01
4.68
1.40
.00
2.11
1.82
.25
4.77
.90
<.0001
Building and
grounds cleaning
and maintenance
Coefficient
Standard error
p–value
1.45
3.30
.66
.71
.91
.44
2.02
1.91
.29
5.12
.72
<.0001
2.91
.58
<.0001
3.06
.89
.00
–.11
1.16
.93
2.23
.57
<.0001
Personal care and
service
Coefficient
Standard error
p–value
2.33
5.45
.67
–3.43
1.50
.02
–7.18
3.16
.02
–10.05
1.19
<.0001
1.76
.95
.07
.89
1.47
.55
.40
1.92
.83
.39
.94
.68
Sales and related
Coefficient
Standard error
p–value
–9.54
4.56
.04
–16.01
1.26
<.0001
–7.41
2.64
.01
–10.39
1.00
<.0001
–4.05
.80
<.0001
–9.25
1.23
<.0001
–7.05
1.60
<.0001
–6.20
.79
<.0001
See note at end of table.
32 Monthly Labor Review • June 2011
Table A–4. Output for regressions of change in employment in each occupational group on all 8 layoff region interaction
variables, and controls
Dependent variable
(change in employment in occupational group)
Output for
interaction
terms
Interaction term
New
England
New York/
New Jersey
Mid–
Atlantic
Southeast
Midwest
Southwest
Mountain–
Plains
West
Office and
administrative
support
Coefficient
Standard error
p-value
–18.65
11.02
.09
–13.33
3.03
<.0001
–29.98
6.39
<.0001
–15.98
2.41
<.0001
–0.07
1.93
.97
–14.21
2.98
<.0001
–12.41
3.88
.00
–15.12
1.91
<.0001
Construction and
extraction
Coefficient
Standard error
p-value
.80
3.88
.84
–2.49
1.07
.02
–2.62
2.25
.25
.27
.85
.75
–4.14
.68
<.0001
–2.32
1.05
.03
–3.54
1.37
.01
–.16
.67
.81
Installation,
maintenance, and
repair
Coefficient
Standard error
p-value
–8.15
4.09
.05
–1.19
1.13
.29
–7.78
2.37
.00
–10.40
.89
<.0001
–3.12
.72
<.0001
9.23
1.10
<.0001
–4.77
1.44
.00
–4.52
.71
<.0001
Production
Coefficient
Standard error
p-value
–29.86
8.90
.00
–7.14
2.45
.00
–53.37
5.16
<.0001
–48.97
1.95
<.0001
–23.78
1.56
<.0001
–.55
2.40
.82
–42.18
3.13
<.0001
–6.93
1.54
<.0001
Transportation and
material moving
Coefficient
Standard error
p-value
1.66
7.03
.81
–12.61
1.94
<.0001
–15.27
4.08
.00
–14.35
1.54
<.0001
–4.61
1.23
.00
–4.59
1.90
.02
7.54
2.48
.00
–3.95
1.22
.00
NOTE: Table excludes output for control variables to conserve space, and excludes farming, fishing, and forestry occupations. Degrees of freedom=1 for
all regressions.
Monthly Labor Review • June 2011 33